Metaphor and Iconicity: A Cognitive Approach to Analysing Texts

Metaphor and Iconicity A Cognitive Approach to Analysing Texts Masako K. Hiraga Metaphor and Iconicity Also by Masa...

Author: Masako K. Hiraga

77 downloads 1359 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Metaphor and Iconicity A Cognitive Approach to Analysing Texts

Masako K. Hiraga

Metaphor and Iconicity

Also by Masako K. Hiraga CULTURAL, PSYCHOLOGICAL AND TYPOLOGICAL ISSUES IN COGNITIVE LINGUISTICS (Co-editor with Chris Sinha and Sherman Wilcox)

Metaphor and Iconicity A Cognitive Approach to Analysing Texts Masako K. Hiraga

© Masako K. Hiraga 2005 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP. Any person who does any unauthorised act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The author has asserted her right to be identiﬁed as the author of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2005 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N.Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN 1–4039–3345–6 hardback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Hiraga, Masako. Metaphor and iconicity : a cognitive approach to analyzing texts / Masako K. Hiraga. p. cm. Includes bibliographical references and index. ISBN 1–4039–3345–6 (cloth) 1. Metaphor. 2. Iconicity (Linguistics). 3. Discourse analysis. 4. Poetics. 5. Cognitive grammar. I. Title. P301.5.M48H57 401¢.43–dc22

2004 2004052323

10 9 8 7 14 13 12 11

6 10

5 09

4 08

3 07

2 06

1 05

Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham and Eastbourne

Contents Acknowledgements

ix

List of Tables

xiii

List of Figures

xiv

List of Abbreviations

xvi

Part I 1

2

Framework

1

Introduction 1.1 A glimpse of the issue 1.1.1 Metaphor in icon 1.1.2 Iconicity in metaphor 1.1.3 How a poem manifests metaphor–icon links 1.2 Purpose of study 1.2.1 Contribution to the study of linguistic iconicity 1.2.2 Contribution to the study of the status of written language 1.3 Scope and organisation of study 1.3.1 Scope of study 1.3.2 Organisation of study Methodological Framework 2.1 Background literature 2.1.1 The traditional treatment of metaphor and iconicity 2.1.2 Metaphor and iconicity in cognitive linguistics and poetics 2.2 Peircean deﬁnition 2.2.1 Firstness, Secondness, Thirdness 2.2.2 Icon, index, symbol 2.2.3 Image, diagram, metaphor 2.3 Cognitive deﬁnition: interplay between metaphor and iconicity 2.3.1 Iconicity and metaphor deﬁned 2.3.2 The model of blending v

3 3 3 5 7 14 14 18 19 19 20 22 23 23 25 29 29 30 31 35 35 36

vi

Contents

2.4

2.5 Part II 3

2.3.3 Grammatical metaphors Jakobson on poetic function revisited 2.4.1 The poetic function of language 2.4.2 The role of repetition in text 2.4.3 Repetition as grammatical metaphor Summary

41 45 46 46 49 50

Analysis

55

Manifestation of Metaphor–Icon Links: Prototypical Examples 3.1 Shape and sense in Herbert’s ‘Easter Wings’ 3.1.1 Text as shape 3.1.2 Text as icon 3.2 Grammatical parallelism in Shelley’s ‘Love’s Philosophy’ 3.2.1 Text as metaphor 3.2.2 Iconic interpretation by metaphor 3.3 Isomorphism of text structure in Basho’s Oku no Hosomichi 3.3.1 Haiku and cognitive metaphor 3.3.2 Texts 3.3.3 The spring poem 3.3.4 The autumn poem 3.3.5 Metaphor–icon links in macro-structure 3.4 Concluding remarks

57 58 58 59 63 63 67 70 70 72 73 80 85 89

4

Vision as Meaning 4.1 Poetry and visual representation 4.1.1 The writing systems of English and Japanese 4.1.2 Visual representation in poetry 4.2 Visual poetry 4.2.1 Imagic manifestations 4.2.2 Diagrammatic manifestations 4.2.3 Remarks on comparison 4.3 Subtle manifestations of visual iconicity in poetry 4.3.1 General diagrammatic features 4.3.2 Orthographical revisions of haiku 4.4 Concluding remarks

91 91 92 96 100 100 106 115 116 117 118 125

5

Sound as Meaning 5.1 Poetry and auditory representation 5.1.1 The sound systems of English and Japanese 5.1.2 Auditory representation in poetry

127 127 128 131

Contents vii

5.2

5.3

5.4 Part III 6

7

8

Phonosymbolism 5.2.1 The dance of sounds in Edgar Allan Poe’s ‘The Bells’ 5.2.2 Charging the music: phonological revisions of haiku Diagrammatic structuring of the sound 5.3.1 Mirror of sounds in Williams’s ‘The Red Wheelbarrow’ 5.3.2 Hidden melodies in Tomonori’s ‘Hisakata no’ Concluding remarks Further Issues

The Interplay of Metaphor and Iconicity in Grammar and Discourse 6.1 A typology of grammatical metaphors 6.1.1 Relational diagrams 6.1.2 Structural diagrams 6.1.3 Iconicity in discourse 6.2 Metaphor–icon links in polite language 6.2.1 VERTICAL DISTANCE 6.2.2 TEMPORAL DISTANCE 6.3 Concluding remarks

134 135 146 156 157 161 168 171 173 173 175 178 185 186 187 191 192

Iconicity and Written Language 7.1 Written language and writing systems 7.1.1 The alphabet vs non-alphabetic writing systems 7.1.2 Chinese characters as ideograms and logographs 7.2 Iconicity and kanji 7.2.1 Metaphor–icon links in kanji 7.2.2 Kanji as a poetic medium 7.3 A re-evaluation of written language 7.3.1 The comprehensive view 7.3.2 An extended deﬁnition of written language

194 194

197 199 200 206 207 207 212

Conclusions 8.1 Summary of main claims 8.1.1 Theoretical contributions 8.1.2 Methodological contributions 8.2 Further research issues

219 219 219 222 224

195

viii Contents

8.2.1 8.2.2

‘Languaging’: language in the making Ecology of communication media

224 225

Notes

228

References

239

Index

253

Acknowledgements

I would like to acknowledge the many people who have helped and encouraged me both before and during the writing of this book.

Central to this project are the vision and insight of Mark Turner and Gilles Fauconnier, who deserve my greatest debt of gratitude. I would like to thank them for their works and words, which always boosted my conﬁdence and sharpened my thinking.

On the entire manuscript, Hayley Davis and Joan Turner offered extensive comments and suggestions. As friends and colleagues, they have meant so much to me. I would also like to express my gratitude to Yoshihiko Ikegami, Gunther Kress and Michael Toolan for their constructive criticisms on early drafts of the manuscript. Not only for being my ﬁrst mentor in linguistics at Aoyama Gakuin University, Tokyo, but also for his continued inﬂuence through the research circle of the Institute of Human Linguistics (IHL), I must thank Yasuo Isami. Members of IHL who have helped through their various discussions include Masako Akimatsu, Haruo Hoshii, Katsutoshi Ito, Masaru Makiuchi, Yumiko Mikuni and Niro Ohki.

Ever since we ﬁrst met, Yoko Fujii, Joanna Radwanska-Williams and Valerie Anne Wilkinson have been constant friends and the most ideal of colleagues. Without their rich input at every stage of thought and writing, the book would have been much the poorer.

The moral support, encouragement and friendship of Erich Berendt, Brian Falconbridge, Masanori Higa, Sachiko Ide, Teruyuki Kume, Kenichi Noda, Kumiko Torikai and Yasukata Yano are highly appreciated.

A

special thank you goes to Joseph Goguen, John Robert Ross and Linda Waugh, whose feedback has been invaluable in shaping my ideas and challenging my thought. ix

x Acknowledgements

Principal among many other linguists to whom I have referred in the book, are several close friends and colleagues who deserve special thanks, including Michiko Arima, Donald Freeman, Margaret Freeman, Hartmut Haberland, Jacob Mey and Eve Sweetser. Home away from home in England was provided by Dorrik Stow, Lynn Watson and their families. I also appreciate detailed editorial assistance given by Dorrik Stow.

Of all my friends and colleagues, I wish to thank Rosalia Dutra, Ranjana Goel, Momoko Hirayama, Hiroko Ito, Akiko Kawasaki, Junko Nagato, Emiko Noda, Akiko Sakuma, Kyoko Sakuma, Michiyo Somura, Suk-Ying Wong and Naoko Yazawa, for their compassion and support. My thanks also go to Keiko Naruoka, Ayako Shibata, Chiho Sunakawa, Hiroko Takanashi, Kyoko Watanabe and Tomoko Watanabe, for their assistance at various stages of this project. I am especially grateful to Jill Lake, my editor at Palgrave Macmillan, for her guidance and insight. Real gratitude throughout my life goes to my late parents, Kimpei and Yoshiko Koshiyama. I would also like to thank Shigetaka and Tomoko Hiraga, for their support and encouragement. This book is dedicated to my husband, Nobutaka, who knows how much I must thank him.

Acknowledgements xi

Acknowledgements for material in the text Sections 1.1.1, 1.1.2, 3.1 and 3.2 are reprinted from ‘The Interplay of Metaphor and Iconicity: a Cognitive Approach to Poetic Texts’. Special Issue on Metaphor, edited by Tomasz Komendzinski, Theoria et Historia Scientiarum: International Journal for Interdisciplinary Studies, VI/1 (2002), pp. 179–244. Reprinted with the permission of Uniwersytet Mikolaja Kopernika, Poland. Table 2.2 is reproduced with a slight alteration from ‘Diagrams and Metaphors: Iconic Aspects in Language’. Special Issue on Metaphor and Iconicity, edited by Masako K. Hiraga and Joanna Radwanska-Williams, Journal of Pragmatics, 22(1) (1994), pp. 5–21. Reproduced by permission of Elsevier. Figure 2.1 is reproduced with a minor alteration from the blending website at http://markturner.org/blending.html, with the permission of Gilles Fauconnier and Mark Turner. Figure 2.5 is reproduced with a slight alteration from ‘How Metaphor and Iconicity Are Entwined in Poetry: a Case in Haiku’, in Wolfgang G. Müller and Olga Fischer, From Sign to Signing (2003), pp. 317–35. With kind permission by John Benjamins Publishing Company, Amsterdam/ Philadelphia. www.benjamins.com ‘l(a’ is reprinted from Complete Poems 1904–1962, by E. E. Cummings, edited by George J. Firmage, by permission of W. W. Norton & Company. Copyright © 1991 by the Trustees for the E. E. Cummings Trust and George James Firmage. ‘Swan and Shadow’ is reprinted from Types of Shape, by John Hollander, by permission of Yale University Press. Copyright © 1969, 1991 by John Hollander. ‘40-Love’ is from After the Merrymaking, by Roger McGough, published by Jonathan Cape, London. Reprinted by permission of PFD on behalf of Roger McGough. Copyright © 1971 Roger McGough. ‘The Red Wheelbarrow’ by William Carlos Williams, from Collected Poems: 1909–1939, Volume I, copyright © 1938 by New Directions Publishing Corp. Used by permission of New Directions Publishing Corporation.

xii Acknowledgements

‘In a Station of the Metro’ is from Collected Shorter Poems by Ezra Pound, copyright © 1928. Reprinted by permission of the author and Faber and Faber Ltd. Tables 5.5, 5.6, 5.8, and Figures 5.7 and 5.8 are reproduced from ‘Eternal Stillness: a Linguistic Journey to Basho’s Haiku about the Cicada’. Poetics Today, 8(1) (1987), pp. 5–8. Reproduced by permission of Duke University Press. Figure 6.3 is reproduced from ‘Deference as Distance: Metaphorical Base of Honoriﬁc Verb Construction in Japanese’, in Masako K. Hiraga, Chris Sinha and Sherman Wilcox (eds), Cultural, Psychological and Typological Issues in Cognitive Linguistics (1999) pp. 47–68, by kind permission of John Benjamins Publishing Company, Amsterdam/Philadelphia. www.benjamins.com Figure 7.1 Cretan pictographic inscriptions in early Minoan Period. Extract from The Alphabet: a Key to the History of Mankind by David Diringer published by Hutchinson. Used by permission of The Random House Group Limited. Every effort has been made to trace all copyright holders, but if any have been inadvertently overlooked, the publisher will be pleased to make the necessary arrangements at the ﬁrst opportunity.

List of Tables 1.1 1.2 2.1 2.2 2.3 2.4 2.5 3.1 3.2 4.1 4.2 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11

Distribution of vowels Distribution of consonants Icon, index and symbol Subtypes of icon Grammatical metaphors Units of repetition Iconic meanings of repetition Lexical pairs in ‘Love’s Philosophy’ Stanzaic pairs in ‘Love’s Philosophy’ Characteristics of character types Different types of revisions Onomatopoeic words Sound-symbolic elements Alliteration and rhyme in the framing lines Repeated lexical items Revision of the verbs Revision in the vowels Occurrence of vowels in the 50 haiku in Oku no Hosomichi Sound pattern of vowels in sequence Mirror-image of vowels Mirror-image of consonants Number of occurrence of consonants

xiii

12 12 30 34 42 47 49 69 69 93 119 140 140 143 144 149 150 150 151 160 160 164

List of Figures

2.1 2.2 2.3 2.4 2.5 2.6 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.1 4.2 4.3 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 6.1 6.2 6.3 6.4

The model of blending Metaphor–icon links in the model of blending Structural diagram Relational diagram Manifestations of repetition types Metaphor–icon links in language Metaphor–icon links in ‘Easter Wings’ Metaphor–icon links in ‘Love’s Philosophy’ Blend of the spring poem Metaphor–icon links in tori naki Metaphor–icon links in futami Revision from tachikaeru to yukikau Use of the logograph in the paralleled texts Use of the logograph in the beginning and the conclusion of the travel sketch Use of the logograph in the macro-structure of text Metaphor–icon links in ‘Swan and Shadow’ Metaphor–icon links in ‘40–Love’ Metaphor–icon links in concrete poetry Metaphor–icon links in ‘The Bells’ Syntactic framing format Phonological and semantic progressions of the text Oneness of silence and voice Metaphor–icon links in ‘Shizukasa ya’ Mirror-image of syllables Metaphor–icon links in ‘The Red Wheelbarrow’ Distribution of [h] and [n] Distribution of [k] and [r] Kokoro inside hana Metaphor–icon links in ‘Hisakata no’ Relational diagram Structural diagram

37 40 44 45 48 51 60 67 74 79 84 87 88

DEFERENCE IS VERTICAL DISTANCE

88 89 104 107 108 139 143 152 153 154 159 161 164 165 167 168 176 179 189

Metaphor–icon links in grammaticalisation of verbs of vertical motion

190

xiv

List of Figures xv

7.1 7.2 7.3 7.4 7.5

Cretan pictographic inscriptions in the early Minoan period Modern ideograms Shookei-moji of ‘gate’ and ‘the moon’ Metaphor–icon links in Metaphor–icon links in

196 197 201 202 203

List of Abbreviations ACC accusative ADJ adjective AUX auxiliary DAT dative GEN genitive GER gerundive GOAL goal HON honoriﬁc IMP imperative INT intensiﬁer N noun NEG negative PAST past POL polite PRES present QUES question SUB subject TAG tag TOP topic

xvi

For Nobutaka

This page intentionally left blank

Part I Framework

This page intentionally left blank

1 Introduction

The ubiquity and mutual implication of Verb and Verbal Art impart a seminal unity to the forthcoming science of the two inseparable universals, Language and Poetry. Roman Jakobson and Linda R. Waugh (1979) This book attempts to clarify the interplay of metaphor and iconicity in the creation and interpretation of spoken and written discourse. The claim is that the interplay is particularly foregrounded in the case of poetic texts. By analysing a number of short Japanese and English poems, I will argue against the dominant view of language that sees the linguistic sign as primarily non-iconic or arbitrary. This chapter offers an overview of the issues to be discussed in this study.

1.1 A glimpse of the issue 1.1.1 Metaphor in icon Icons resemble objects It is only recently that the word ‘icon’1 has become readily understood as a loan word in Japanese. This is because of the pervasiveness of computer culture. When we actually manipulate the ‘folders’, ‘documents’ and ‘wastebasket’ on the ‘desktop’ of our computer screen, we can see the shape of the folders and open, move, close or even discard them into the wastebasket as we wish.2 We can understand what the icons stand for even without instructions because the folders and documents on the screen resemble the objects that we already know in our ofﬁce. The folder on the screen is an icon for a folder, the document for a document, the wastebasket for a wastebasket, the desktop (the term 3

4

Metaphor and Iconicity

actually refers to the screen itself) for the surface of a desk. Icons resemble the objects they stand for in terms of their shape, form, appearance or structure. Icons are easy to understand because their resemblance to the objects is immediate and concrete. A moment’s reﬂection, however, makes us realise that the folder, document, wastebasket and desktop are very different from the objects that we actually have in our ofﬁce. Folders in our ofﬁce do not include other folders in them, while folders on the screen can, in theory, contain an unlimited number of other folders. In our ofﬁce, we do not usually put a wastebasket on the top of our desk, but the desktop of the computer screen has a wastebasket on the same surface as the ﬁle-folders.3 The metaphorical navigation of icons The use of the words ‘folder’ and ‘desktop’ to signify these icons is metaphorical in a technical sense, because we are making connections between different things based on certain similarities. The folders represented by icons on the screen as well as the folder objects in our ofﬁce are containers of information used for the purpose of storage and classiﬁcation. The desktop is a place where we work both on the screen and in the ofﬁce. When we call the rectangular signs on the screen folders and documents, the names for the icons are a metaphorical extension of the original items, which signify the referents (i.e. the folder object and the document object). Names such as folder, document and desktop, with their extended metaphorical meaning, cue the interpretation of the iconic signs. If we call these iconic signs by different names, say, ‘box’ or ‘rectangle’, it would be difﬁcult to draw the connection, because the metaphorical extension is less obvious. In short, a metaphor reinforces or navigates the iconic meaning of the sign. Iconicity dominance in metaphor–icon links The folders, documents and wastebasket on the computer screen can be described as both icons and metaphors at the same time. They are icons because they look similar to the objects they represent. They are metaphors because, through metaphorical extension of the names given, certain properties of the concept of folder, document and wastebasket are cognitively mapped onto the signs on the screen, or onto the mathematical computation represented by those signs. Because of the visual resemblance of the folder and wastebasket icons with actual folders and wastebaskets, their iconicity is more readily noticeable than their metaphorical status. In other words, while iconicity is more

Introduction 5

dominant than metaphor in the computer screen example, both iconicity and metaphor are at work in the signiﬁcation of highly iconic signs. 1.1.2 Iconicity in metaphor Metaphorical understanding Metaphor and iconicity occur together at the level of metaphorical signiﬁcation, too. As computer jargon became popular, we began to use it when thinking and talking about something else. For example, when we say, ‘I have too many ﬁles open in my mind, and my mind is jammed’, we are referring to our mind as if it were a computer.4 Here, we are not only using expressions from computer vocabulary, but we are also conceptualising our mind as if it were a computer.5 So, we also say the expressions given in Example 1.1. Example 1.1 (a) My system crashed. (b) He has a read-only memory. (c) You gave me a lot of input. (d) I have too much on my desktop. (e) She’s in emulation mode. Metaphor allows us to understand a relatively abstract and unstructured subject matter (technically called a target domain) in terms of a more concrete and structured subject matter (a source domain). In our example, we use the terms designating computer processes as tools to represent the workings of our own minds. We know how to install, save, and delete information in the computer because these are the terms that describe our practical experience with the computer. Therefore, it is natural to apply the pattern knowledge gained from the experience of using a computer to the invisible, complex functioning of our own brain. This is the experiential basis for metaphorical conceptualisation. The image content of metaphor The same process that made computer icons of the ﬁles, desktops and wastebaskets of the conventional ofﬁce makes those same icons useful for conceptualising the functioning of the brain; we use items which are cognitively accessible as the source for metaphors. Metaphors may be imaged from visual, auditory and other sensory experience. When we say, ‘I have too many ﬁles open in my mind’, it is easy to evoke the image of a screen with many ﬁles open. And the image evoked in our

6

Metaphor and Iconicity

mind is similar to the actual computer screen with many ﬁles open. In this way, we can say that the image content of metaphor, particularly of a source domain, is an iconic moment involved in metaphor. In Danesi’s words (1995: 266), ‘the particular content of a metaphor can be said to constitute an interpretation of reality in terms of mental icons that literally allows us to see what is being talked about’ (italics in the original). The diagrammatic content of metaphor Not only the image content but also the correspondence of the two items mediated by metaphor signal an iconic moment. As mentioned earlier, metaphor is a mapping from a source to a target. Each mapping consists of a set of correspondences between entities in the two domains. In our example, some of the characteristics of computers are mapped onto what we know about the human mind. For example, ﬁles correspond with the storage of ideas, crash with no functioning, readonly memory with a lack of learnability, input with information, and emulation mode with imitating. Technically speaking, this mapping is a projection of image-schemas from the source domain onto the target domain. Image-schemas are skeletal patterns or structures that recur in our sensory and motor experiences, perceptual interactions and cognitive operations (cf. Johnson 1987). For example, when we say, ‘I have too many ﬁles open in my mind’, the image-schemas of excess in the source (‘too many opened ﬁles’) and in the target (‘too many ideas’) are aligned. Note that there is an iconic moment in this alignment of image-schemas. The image of too many ﬁles corresponds to the image of too many ideas diagrammatically. In other words, they resemble each other in terms of the image-schemas of excess. There is a structural analogy between the two image-schemas. Metaphor dominance in metaphor–icon links While iconicity is more dominant in the computer screen example, metaphor is more dominant than iconicity in the signiﬁcation process of MIND IS A COMPUTER, because the projection of the COMPUTER domain onto the HUMAN MIND domain is more conceptual than perceptual (visual, auditory or sensory). Nonetheless, the experiential motorsensory image base and schematic parallel structure are iconic moments in the metaphorical signiﬁcation. It may not be easy to recognise the interplay of metaphor–icon links in the above examples, because the interplay is backgrounded by the dominance of metaphor over iconicity or of iconicity over metaphor in

Introduction 7

their manifestations. However, a different picture emerges once attention is paid to foregrounded cases such as poetic texts or signed languages.6

1.1.3 How a poem manifests metaphor–icon links In this section, a haiku text composed by Basho Matsuo (1644–94) is considered as an illustration of metaphor–icon links in poetry. Before going on to analyse the haiku in detail, it is necessary to give some background information on haiku as a poetic form. Haiku, or hokku as it was formerly called during the time of Basho, is the shortest form of Japanese traditional poetry. It comprises 17 morae,7 divided into three sections of 5–7–5. Haiku originated from the ﬁrst three lines of the 31-mora tanka, and began to rival the older form in the Edo period (1603–1867), evolving into a more serious and profound art form as practised by the great master Basho. It has since become and remained the most popular poetic form in Japan. At ﬁrst, the subject matter of haiku was restricted to a simple description of nature, with veiled reference to one of the seasons, and was intended to evoke a clear emotional response in the reader. Later, its subject range was broadened considerably but was always striving to express as much as possible in the fewest possible words. From very early in their history, both tanka and haiku have been and still are composed by people of every class, men and women, young and old. There is no attempt to achieve any form of rhyming, in the sense of European poetry, as the Japanese language has only ﬁve vowel sounds, [a], [e], [i], [o] and [u], with which to form its morae, either by themselves or in combination with a consonant as in consonant–vowel sequences. Brevity, suggestiveness and ellipsis are the ‘life and soul’ of these poetic forms, and much is left for the reader’s own imagination.8 The haiku to be looked at here is taken from Basho’s Oku no Hosomichi, one of the acknowledged masterpieces of Japanese literature (Example 1.2). Example 1.2 araumi ya rough sea:

Sado ni yokotou Sado in lie

ama no gawa heaven of river9

‘Rough sea: lying toward Sado Island the River of Heaven’ (Matsuo 1996 [1694]: 109).

8

Metaphor and Iconicity

Metaphorical juxtaposition At ﬁrst reading, the poem describes a natural scene with two contrasting elements – a rough sea lying between the poet and Sado Island, and the Milky Way (River of Heaven) overhead. Even without any pragmatic knowledge about Sado Island or the Milky Way in Japanese history and culture, it is possible to sense something of the grandeur and even the danger that this haiku evokes with such eloquence. It is a starry night with an awe-inspiring Milky Way arching across the sky towards Sado Island. The peaceful, silent grandeur of the Milky Way is placed in stark contrast to the rough, noisy sea that churns and moans, so dangerous and terrifying in the darkness. Indeed, the arm of the Sea of Japan that separates Sado Island from the mainland can be wild and rough and has taken countless lives of ﬁshermen and island residents over the years. Though the island is clearly visible across the troubled waves, perhaps with scattered houselights in the distance, the passage is often dangerous especially at night. Human beings (including the poet) are so small and helpless in the face of this spectacular pageant of powerful nature. The individual stories of many thousands of human lives are etched forever in the scene, almost as stars across the Milky Way. The ﬁrst line or segment, araumi ya, consists of a noun, araumi (‘rough sea’), and a kireji (‘cutting letter’), ya. There are about a dozen particles, or kireji, typically used in tanka and haiku, that mark a division point in the text and leave room for reﬂection on the feelings or images evoked by the preceding segment. The ya in this example is a kireji particularly favoured by Basho and said to have ‘something of the effect of a preceding “Lo!” It divides a haiku into two parts and is usually followed by a description or comparison, sometimes an illustration of the feeling evoked. There is always at least the suggestion of a kind of equation, so that the effect of ya is often best indicated by a colon’ (Henderson 1958: 189). In this case, araumi (‘rough sea’) is metaphorically juxtaposed against the rest of the text. The next line, Sado ni yokotou (‘[which] lies toward Sado’), is an adjectival clause which modiﬁes the last segment, ama no gawa (‘the river of heaven’). Sado is the name of an island located about 50 miles southeast of the mainland, ni (‘toward’) is a postpositional particle of location, and yokotou (‘to lie’) is a verb that describes the action of spreading one’s body on something ﬂat (when used as a transitive verb) or the state of a body spread out on something ﬂat (when used as an intran-

Introduction 9

sitive verb). It normally has an animate agent, but as the grammatical subject in this poem, ama no gawa (‘the river of heaven’), is an inanimate noun, the verb is used metaphorically. The last line, ama no gawa, is a proper noun literally translated as the river of heaven, but here signifying the Milky Way. In itself it is a metaphor in which the broad band of stars (the Milky Way) is seen as a river. The second and the third segments of the poem thus constitute a local metaphor, in which a river of stars in the night sky spreads its body toward Sado Island. The conventional metaphors behind this local metaphor are that NATURE IS ANIMATE (in this case RIVER IS ANIMATE)10 and that A PATH-SHAPED OBJECT IS A RIVER. In addition to understanding the grammatical structure and convention used in the poem, further insight is gained by considering the historical setting and cultural tradition at the time of writing. Firstly, the Island of Sado has a rich and colourful past. It is separated from the mainland by a narrow but hazardous arm of the Sea of Japan. Because boat crossings were especially fraught with danger, the island was used as a place of exile for felons and traitors from the tenth to late nineteenth century. But the discovery of gold on the island, in the early seventeenth century, attracted fortune seekers as well as the already wealthy. At the time of Basho, the Tokugawa Shogunate had control of the gold mines, and the people imprisoned on the island were forced to serve as free slave labour. The metonymy of a rough sea with Sado Island evokes its cultural heritage, while the roughness of the waves is consonant with an undoubtedly harsh life on the island, one in which violence, cruelty and despair were rife. Secondly, there is a sad legend about the Milky Way, which originated in China and was brought to Japan. On either side of the Milky Way, there are two particularly bright stars, Vega and Altair, which are taken to represent Princess Weaver and Oxherd. Although they face each other every night, the Milky Way is so wide they cannot easily meet, so one day the god of heaven, pitying Princess Weaver and her lonely life, arranged for her to marry Oxherd. Soon after they married, the Princess became too lazy to weave, so the god became angry and punished her. She was to be allowed to visit her husband only once a year, on the seventh night of the seventh month of the lunar calendar, but only if the night was fair. The date on which this poem was composed (6th July) suggests that the poet had this legend in his mind. For the seventh night of the seventh month (7th July) is known and celebrated as the ‘star festival’

10

Metaphor and Iconicity

after the Chinese story. Basho has cleverly used the legendary separation of this couple as a hidden metaphor for the people imprisoned on Sado Island away from their loved ones at home. Both the Milky Way and the Island of Sado, therefore, share event frames for conﬁnement – spatial conﬁnement, limited freedom, limited means of travel and the mental state of being separated. A third important point lies in the name of the island when written in Chinese kanji. It is made up of two Chinese logographs, and , which mean ‘to help’ and ‘to cross’ respectively. The cognitive meanings of these logographs are both intricately bound into the fabric of the poem. It is also possible to detect, in the alignment of the two imageschemas, rough sea and island, the use of such salient conventional metaphors as LIFE IS A BOAT JOURNEY and THE WAVES ARE AN OBSTACLE TO SUCH A JOURNEY. In summary, this most exquisite haiku of Basho comprises a multitude of metaphors in which numerous images are overlain and integrated to produce the multi-layered meanings that become apparent on careful reading. Such images include the rough sea and Milky Way as obstacles, but also as potential bridges toward union, the legendary couple and prisoners as people prevented from reuniting with their loved ones. Furthermore, there are other more deeply hidden images, together with the feelings they evoke – for example, of sorrow at separation, or a realisation of the helplessness of human beings in the awesome face of nature. There is ever an implicit contrast – for example, a contrast of motion between the violent waves and the peaceful skies; a contrast of light between the dark sea and the starlit Milky Way; and a contrast of reality and legend between life stories of people and the love story of stars. Iconic interpretation navigated by metaphor The true richness of Basho’s poem lies not only in the multiple layers of metaphor, but also in the visual and the auditory conﬁguration of the text, which reveals a remarkable iconicity. In terms of the visual elements, the Japanese language has a unique writing system in which three different types of signs are used to describe the same phonological text: kanji (Chinese logographs), hiragana (syllabary for words of Japanese origin) and katakana (syllabary for words of foreign origin other than Chinese). The choice and use of particular logographs are of particular importance because they function as a cognitive medium for poetry. Basho was, of course, well aware of this fact, which led him to make an orthographic revision of an earlier

Introduction 11

version of the poem from Example 1.3 (a) to 1.3 (b) (Matsuo 1957 [1694]). Example 1.3 (a) (b) araumi ya Sado ni yokotou ama no gawa The three noun phrases, araumi, Sado and ama no gawa, were all written in kanji in both the ﬁrst (Example 1.3 (a)) and the revised (Example 1.3 (b)) versions. However, the verb yokotou (‘to lie’) was changed from kanji to hiragana (see boxed sections). This had the effect of making that part of the text a ground for the conspicuous proﬁle of (‘rough sea’) and (‘Milky Way’). In general, because kanji have a distinct angular form and semantic integrity, they differentiate themselves visually and cognitively as the ﬁgure while the remaining hiragana function as the ground. This differentiation of kanji is particularly prominent in the words, (‘rough sea’), (‘Island of Sado’) and (‘Milky Way’), which are all written in two kanji. Furthermore, all of them include kanji (underlined in Example 1.3(a) and (b)) such as (‘sea’), (‘to cross water’) and (‘river’) which are made up with the same radical signifying water. Both (‘rough sea’) and (‘Milky Way’) relate to water, as described above. The semantic similarity between (‘rough sea’) and (‘Milky Way’) in terms of wateriness and the obstacle (in real life and in the legend explained above) and their dissimilarity (violence in the ‘rough sea’ and peacefulness in ‘the river of heaven’) are also foregrounded. This is a case of diagrammatic iconic effect, whereby the meaning of the foregrounded elements is intensiﬁed by the repetitive use of similar visual elements – two-character nouns and the same radical. The logograph in Sado , the name of the island, is also important because it means ‘to cross water’. As both history and legend show, the ‘rough sea’ and ‘the Milky Way’ are each obstacles that the loved ones must cross in order to meet. This character is aptly placed in the very middle of the poem as if it signalled the crossing point. The sound structure also exhibits interesting iconic effects, which are navigated by the metaphorical interpretation. The following analysis illustrates three possible iconic effects produced by the distribution of vowels, consonants and the repetition of adjacent vowels. Example 1.4 is a phonological notation of the poem’s moraic structure.

12

Metaphor and Iconicity

Example 1.4 Line 1 a-ra-u-mi ya ([-] = the division between mora) Line 2 sa-do ni yo-ko-to-u Line 3 a-ma no Na-wa Firstly, the nature and distribution of vowels show a predominance of back vowels such as [a] and [o]. As indicated in Table 1.1, there are 8 [a] (47 per cent) and 5 [o] (29 per cent) out of 17 vowels.11 Both vowels [a] and [o] are pronounced with a wide passage between the tongue and the roof of the mouth, and with the back of the tongue higher than the front. The backness and the openness of these vowel sounds can create a sonorous effect, which in turn may draw associations of something deep and large (cf. Jespersen 1964 [1921]). In Basho’s poem, these sounds may evoke something of the size of waves, the wide expanse of dark sea, and the broad width of the Milky Way. Secondly, sonorous effects are also created by the frequent use of nasals ([m], [n], and [N]), and vowel-like consonants ([y] and [w]). Table 1.2 shows the nature and distribution of consonants. The dominance of sonorants such as [m], [n], [N], [r], [y] and [w] is clearly evident. They provide prolongation and fullness of the sounds, and hence usually produce lingering effects (cf. Shapiro and Beum 1965: Table 1.1 Distribution of vowels a

o

i

u

e

Total

Line 1 Line 2 Line 3

3 1 4

0 4 1

1 1 0

1 1 0

0 0 0

5 7 5

Total

8

5

2

2

0

17

Table 1.2 Distribution of consonants Position of mora in line

Line 1 Line 2 Line 3 Total

1

2

3

4

5

6

s

r d m

n n

m y N

y k w

t

No. of sonorants

No. of obstruents

3 2 4

0 4 0

9

4

7

Introduction 13

10–12). As such, it could be argued that the back vowels and sonorant consonants jointly reinforce a sound-iconic effect of the depth, breadth and large size of water elements in the poem. It is also signiﬁcant that the only line that has obstruents (i.e. non-sonorants such as [s], [d], [k] and [t]), is Line 2, in which the island is mentioned. If we interpret sonorants as iconically associated with water elements, then obstruents are associated with non-water, namely, the island in this text. Thirdly, the text seems to conceal very cleverly and, perhaps wittingly, a keyword, which is congruous with the metaphorical interpretation of the poem. The prototypical sound sequence in Japanese is an alternation of a single consonant and a single vowel such as CV–CV–CV. This general feature applies to Basho’s haiku under discussion. A closer look, however, reveals just three occurrences of two vowels adjacent to each other – [a–u], [o–u] and [a–w], since [w] is phonetically close to [u]. They occur in araumi in Line 1, yokotou in Line 2, and as the similar sound sequence, gawa, in Line 3. Each line of the poem, therefore, has a similar vowel sequence hidden within the text. Very interestingly, these vowel sequences, [a–u] and [o–u], are both verbs in Japanese, and , which mean ‘to meet’. Their hidden repetition in each line can be read as the echo of a hidden longing between separated persons. Once again, the iconic effect of this hidden element can only be guided by reading the whole poem as a global metaphorical juxtaposition, i.e. the separation of stars, of people and of loved ones.

Summary Compared to the simpler examples from computer jargon, the poetic text displays the complexity of metaphor–icon links. The basic mechanism of metaphor–icon links, however, is the same, and can be summarised in the two aspects: iconic moments in metaphor and metaphor giving iconic meaning to form. As has been seen, once metaphor–icon links are established in the interpretation of text, such links tend to be taken for granted. Because an iconic interpretation, once apprehended by metaphor, is more immediate and straightforward, it becomes difﬁcult to free oneself from such an interpretation. Moreover, when this process is conventionalised as has been seen in computer jargon, one does not even pay attention to it. It becomes an automatic and therefore an unconscious cognitive operation prevalent in language. It is the aim of this study to attempt to explicate these unconscious cognitive processes, particularly by elaborating the interplay of metaphor and iconicity.

14

Metaphor and Iconicity

Because poetic texts tend to foreground the functioning of these processes, they will be used to illustrate my analysis of the interplay between metaphor and iconicity in the creation and interpretation of spoken and written discourse. In effect, I will argue against the dominant view of language that sees the linguistic sign as being primarily non-iconic or arbitrary. If one is interested in why language is structured in the way it is and why it functions in the way it does, metaphor and iconicity are important issues to be explored in detail and in depth. Iconicity is therefore presented here as a manifestation of the motivated nature of the linguistic sign.

1.2 Purpose of study 1.2.1 Contribution to the study of linguistic iconicity The case against the arbitrariness of the linguistic sign The problem of iconicity has been a recurring theme in the inquiry into the nature of the linguistic sign. Plato’s Cratylus is an early example of the debate about the relationship of form and content in language, between physei (‘nature’) and thesei (‘agreement, convention’) (Plato 1926: 7). In modern linguistics, Saussure asserts that the nature of the association between signiﬁé (‘signiﬁcation’) and signiﬁant (‘signal’) is arbitrary, i.e. not based on any inherent resemblance (Saussure 1983[1916]: 67). Although he sees arbitrariness as a fundamental principle of the linguistic sign, he also admits that the linguistic system itself, the way certain parts of the mass of signs are combined and ordered, is relatively motivated, i.e. arbitrariness is limited. In his Course in General Linguistics (1983 [1916]: 131) Saussure states: Everything having to do with languages as systems needs to be approached, we are convinced, with a view to examining the limitations of arbitrariness. It is an approach which linguists have neglected. But it offers the best possible basis for linguistic studies. For the entire linguistic system is founded upon the irrational principle that the sign is arbitrary. Applied without restriction, this principle would lead to utter chaos. But the mind succeeds in introducing a principle of order and regularity into certain areas of the mass of signs. That is the role of relative motivation. (My italics) Linguists after Saussure have also largely tended to neglect issues relating to motivation.12 One of the ways in which such motivation is manifested is iconicity, in which a form signiﬁes meaning by its resemblance

Introduction 15

to the meaning. There are various degrees in which iconicity manifests itself in language, ranging from prominent manifestations (e.g. onomatopoeia and sound-symbolism) to obscured manifestations (e.g. word order, inﬂectional forms, discourse/text structures, conceptual integration). In order to construct a solid argument, it is essential to establish a valid methodological framework for the concepts used and the data explored. Peircean insight According to Peirce,13 iconicity is divided into three subtypes, i.e. image, diagram and metaphor, based on the degree of abstraction as well as the dominance of characteristics of similarity such as mimicry, analogy and parallelism: Those which partake of simple qualities, or First Firstness, are images; those which represent the relations, mainly dyadic, or so regarded, of the parts of one thing by analogous relations in their own parts, are diagrams; those which represent the representative character of a representamen by representing a parallelism in something else, are metaphors. (Italics in the original) (Peirce 1962 [1955, 1902]: 105) Peircean classiﬁcation of iconicity into three subtypes has not been fully elaborated in the semiotic literature. The notion of Peircean metaphor in particular has received very little attention (Anderson 1984). One possible reason is that Peirce himself did not write much about metaphors as opposed to his lengthy discussion of images and diagrams. The other reason is perhaps that there is a difference in the nature of the similarity relationship among the three subtypes. The ﬁrst two subtypes (image and diagram) deal with a similarity relation between forms and concepts, whereas the third subtype (metaphor) deals with a similarity relation between two concepts. Consequently, those linguistic and poetic studies that have incorporated Peircean iconicity in the scope of their analysis have either discarded issues of metaphor from their studies (Haiman 1985a, b, Waugh 1992, 1994, among others) or treated iconicity as a subordinate issue to metaphor (Lakoff and Johnson 1980, Lakoff and Turner 1989).14 The cognitive approach to metaphor–icon links: a challenge By contrast, this study attempts to clarify the interrelationship of metaphor and iconicity in the context of the cognitive model of metaphor developed by Lakoff, Johnson and Turner (Lakoff and

16

Metaphor and Iconicity

Johnson 1980, Lakoff 1987, 1993, Lakoff and Turner 1989, Johnson 1987, Turner 1987, 1991, among others). Ungerer and Schmid (1996: x) encapsulate the meaning of ‘cognitive’ in their model as ‘an approach to language that is based on our experience of the world and the way we perceive and conceptualise it’.15 Langacker (1998: 1) gives a more concrete deﬁnition by saying that cognitive linguistics emphasises the semiological function of language and the crucial role of conceptualization in social interaction. It [cognitive linguistics] contrasts with formalist approaches by viewing language as an integral facet of cognition (not as a separate ‘module’ or ‘mental faculty’). Insofar as possible, linguistic structure is analyzed in terms of more basic systems and abilities (e.g. perception, attention, categorization) from which it cannot be dissociated. In particular, I would like to draw on Turner and Fauconnier’s work in this theoretical tradition, using the model of ‘blending’16 that they have developed (Turner and Fauconnier 1995a, b, 1999, Turner 1996, 1998, Fauconnier and Turner 1998, 2002). This new model provides a more elaborated and sophisticated account of the mechanism of creativity in metaphor than the earlier Lakoff–Johnson model (Lakoff and Johnson 1980). By introducing four items (two input spaces (a source and a target), a generic space (which includes abstract shared properties of both inputs), and a blended space (in which new meanings are emergent)), the model of blending can specify which part(s) of the metaphorical process relate(s) to the interplay of form and meaning in the text, and thereby can clarify the complexity of systematic creation and interpretation of ambiguity with reference to the ways visual and auditory representation contribute in this process of meaning creation. In cognitive and semiotic terms, icons and metaphors share a common property of signiﬁcation, namely, that ‘motivated’ by similarity. Whether the items to be connected are concrete objects, formal structures, relational properties or abstract concepts, connecting things of similarity is one of the basic operations of the human mind. Iconicity deals with a mapping between form (structure) and meaning in various degrees of abstractness, from concrete attributive similarities between objects to abstract relational analogies between structures of form and meaning. On the other hand, metaphor is a mapping between two conceptual domains of meaning, a projection of a schematised pattern (system) from a less abstract source domain onto a more abstract target domain. The clariﬁcation of metaphor–icon links will strengthen

Introduction 17

the cognitive approach so as to provide a more cohesive and integrated explanation of various linguistic phenomena, including word formation, word order, grammaticalisation, semantic change, poetic discourse, signed languages and writing systems, as manifestations of metaphor– icon links. Data delineation Following Jakobson and Waugh’s assumption (1979) that iconicity is foregrounded in poetic texts rather than in ordinary spoken discourse, my exploration of the interplay of metaphor and iconicity will also be focused on poetic texts. Some texts display pure iconicity. Namely, their visual or auditory form itself mimetically signals the meaning. Other texts show iconicity more subtly through metaphors. For example, a metaphorical reading of the text supports certain iconic patterns prevailing in the textual form such as syntactic structures, lexical choice, phonological patterns, or the choice of characters and letters in writing. Logographic writing systems are particularly conducive to an analysis of iconicity because of their origins, as are the signed languages (cf. Deuchar 1990, Taub 2001). Using Japanese poetic texts, which incorporate logographic writing systems, as principal data for analysis, optimises the challenge of demonstrating the complexity of the interplay between metaphor and iconicity. Consequently, the primary focus of this book is an analysis of the styles of meaning creation by metaphor and iconicity in Japanese short poetic texts, particularly in haiku. Haiku, poetry of 17-mora text-form, is an optimal example of how the human mind connects things with minimum linguistic resources, but in a particularly subtle way.17 Haiku deliberately create rich microcosms, i.e. situations where a multiplicity of interpretations can be apprehended in the poetic text as spoken and written forms. This study argues that metaphor and iconicity are the essence of such mental acts of literary creation, and thus it attempts to analyse their interplay from a cognitive perspective. Since the eighteenth century, there has been an enormous amount of literature on haiku, most of which is philological and palaeographical in nature. A brief list of modern criticism, much of which examines Basho’s haiku, includes Aso (1961), Akabane (1970), Akimoto (1970), Morita (1970), Ogata (1971), Kaneko (1973), Yamamoto (1974), Imoto (1978), Ueno (1986) and Muramatsu (1988). Works on haiku (and tanka) from the perspectives of structural poetics and semiotics include Kawamoto (1978, 1986), Jakobson (1981 [1979]), Ikegami et al. (1983), Kikuchi (1984), Hiraga (1987a, b, 1995a), Arima (1996) and Shirane

18

Metaphor and Iconicity

(1998). It can be said that this study along with Ikegami (1996) and my recent works (Hiraga 1999a, b, 2000a, 2002, 2003) is an attempt to integrate the past philological and structural traditions and reinterpret their methodological contributions in cognitive terms. In this research, therefore, I would like to demonstrate that in language structure and process, and particularly in the poetic texts looked at, a close link between metaphor and icon is manifest in two ways: that there are iconic moments in metaphor; and that a form acquires an iconic interpretation via metaphor. 1.2.2 Contribution to the study of the status of written language Discursive and presentational representation The issue of iconicity involves an examination of writing systems with reference to modes of representation in text. The comparative framework for studying Japanese and English texts undertaken in this study also opens new ground for a discussion of the status of written versus spoken language in a general theory of language. Langer (1957: 79–102) divides the conﬁguration of symbols into two modes of representation: discursive and presentational. The former adds the mode of linear conﬁguration to digital elements, whereas the latter relates to the holistic mode of representation by less discrete elements. Even a quick look at the writing systems of English and Japanese makes one realise that the English alphabet is more discursive than presentational, and kanji or Chinese logographs used in Japanese are more presentational than discursive. The use of three different sign systems (kanji or Chinese logographs, hiragana and katakana or two sets of syllabaries) in Japanese to spell out the same phonetic conﬁguration gives rise to a consideration of linguistic representations from two interrelated perspectives: the written and the spoken language; and discursive and presentational modes of representation. The use of kanji in combination with hiragana and katakana, results in a stronger tendency toward iconicity at the visual and presentational level rather than at the auditory and discursive level. The broader choice of character types makes both poets and readers more conscious of the visual presentation of Japanese poetry. This forms a sharp contrast to the writing systems developed in the Western tradition, in which the alphabetic signs try to record the decomposed sounds of language, and therefore, rhythm and rhyme, the music of poetry, are more essential and prominent (Brooks and Warren 1960: 2). It is no wonder therefore that auditory iconic manifestations are more

Introduction 19

readily recognisable in alphabetic languages, whereas visual iconic manifestations are more readily available in logographic languages. The cognitive role of written language Furthermore, because of their iconic and metaphoric nature, an examination of kanji logographs provides a deeper understanding of the cognitive role of written language in poetic texts. The formation of the shape and the meaning of kanji is seen to be governed by iconic and metaphoric processing. Poetic language therefore exploits these iconic and metaphoric implications of kanji to enrich the complexity and multiplicity of meaning in the text. It is a challenging task of this study to re-evaluate the sharp insight of Fenollosa (Pound 1936) into the nature of Chinese logographs as a medium for poetry and to place it in the new context of cognitive poetics. I believe that a comparative and contrastive study of English and Japanese poetic texts is beneﬁcial since it not only epitomises the commonality of iconic manifestations in diverse texts but also clariﬁes the difference in iconic manifestations in relation to the modes of representation. It also has a bearing on the important but ‘neglected’ issue of the status of written language in a general theory of language.

1.3 Scope and organisation of study 1.3.1 Scope of study By concentrating on the analysis of the interplay of metaphor and iconicity in Japanese haiku and comparable short poems written in English, the study develops a comprehensive methodological programme for the analysis of multiple meanings and interpretations in poetic texts. A consideration of metaphor–icon links will illustrate the fruitfulness of the cognitive approach toward an understanding of poetic discourse. Introducing iconicity and metaphor as an analytical tool has implications within the larger frame of reference of semiotics. To demonstrate such implications, differences in the manifestations of metaphor–icon links will be analysed from the point of view of written versus spoken language with special reference to kanji or logographs used in the Japanese writing systems. Hence, this study, which initially deals with literary texts, ultimately presents a methodology for the examination of iconicity, or a link between form and meaning based on similarity, in language in general.

20

Metaphor and Iconicity

1.3.2 Organisation of study The book is organised in the following way. Part I (Chapters 1–2) provides an overview of the issues to be covered in the study, reviews the basic literature, and offers theoretical and methodological frameworks. Chapter 2 aims to discuss the major background literature and to give the basic framework for analysis. Part II (Chapters 3–5) consists of detailed sample analyses examining how metaphor–icon links manifest themselves on the visual, auditory and conceptual levels in Japanese and English short poetic texts. Chapter 3 provides a comprehensive analysis of how metaphor and iconicity are intertwined with the structure of poetic texts. Four distinctive texts are to be discussed as illustrative cases for further elaboration in the chapters to follow. In Chapter 4, how poetic form embodies meaning in visual terms is explored with respect to the status of written language. Sample analyses of visual poetry will be offered. More subtle manifestations are also analysed with special reference to the orthographical revisions of poems. In Chapter 5, how poetic form embodies meaning by way of auditory manifestation is explored through sample analyses of phonosymbolism and the diagrammatic structuring of sounds in poetry. Part III (Chapters 6–8) summarises the main ﬁndings and discusses their implications for general linguistic issues. In Chapter 6, the interplay of metaphor and iconicity analysed in the poetic texts is reevaluated in the light of general linguistic theory. In particular, metaphor–icon links are looked at as they manifest themselves in grammatical structure and aspects of pragmatic usage. A typology of manifestations of grammatical metaphors is given, and examples of polite language in English and Japanese are analysed with respect to metaphor–icon links. In Chapter 7, the status of written language is considered in relation to the formation and use of kanji logographs. Based on the insights that the logographic writing system provides, the chapter also reconsiders written language as a medium of graphic communication in the era of electronic communication. In Chapter 8, a summary of the main claims is given and further problems are discussed. This book, therefore, contributes to the following three issues: 1 Theoretical clariﬁcation – deﬁning the interrelationship of metaphor and iconicity in the model of ‘blending’; 2 The demonstration of metaphor–icon links in poetic discourse and logographic writing systems – examining how metaphor–icon links

Introduction 21

are manifested in an illustrative analysis of Japanese poetic texts in comparison with English counterparts; 3 The evaluation of the status of written language – providing new ground for discussion on the importance of written language in a general linguistic theory. Thus, this study is a contribution to the development of a theory of poetic language. However, it is also assumed that a detailed survey of the cognitive operations of iconicity and metaphor in poetic texts gives an insight into iconicity and metaphor in language in general,18 and therefore it contributes to a general discussion of the nature of human cognitive processes in linguistic theory.

2 Methodological Framework

Blended spaces do cognitive work in the strongest sense. They provide inferences, emotions, and novel actions, and consequently leave their mark upon the real world. Mark Turner (1996)

Overview of the methodology Methodologically this study draws on three theoretical traditions in the analysis of iconicity and metaphor: Peircean semiotics, the Lakoff–Johnson–Turner cognitive model of metaphor, and Jakobsonian poetics. The study owes its basic insight into the nature of the iconic sign to Peirce (1931–58, 1962[1955, 1902]). He considered metaphor to be one of the three subtypes of iconic signs along with images and diagrams, claiming that metaphor involves the latter two subtypes. While Peirce himself did not particularly elaborate the interrelationship of metaphor and iconicity, it is such an elaboration that this study is concerned with. Speciﬁcally, this interrelationship will be conceived of as two basic manifestations: iconic moments residing in metaphor; and metaphor giving iconic interpretation to form. This study refocuses Peirce’s original conceptualisation with the aid of the cognitive theory of metaphor, which proposes a renewed claim for the motivation of the linguistic sign based on cognitive operations of the mind. In particular the recent theory of ‘blending’ advocated by Turner and Fauconnier (Turner and Fauconnier 1995a, b, Turner 1996, Fauconnier and Turner 1998, 2002, among others) will be developed to clarify the complexity of the interplay of metaphor and iconicity. 22

Methodological Framework 23

The primary data for analysis are taken from poetic texts in which the interplay of metaphor and iconicity is foregrounded. For a detailed analysis of the texts, this study uses the structural approach developed by Jakobson and his advocates (Jakobson 1971, 1981, 1985, Jakobson and Waugh 1979, Jakobson and Rudy 1985 [1980]). Jakobsonian analysis has been criticised on the grounds that it excludes the role of the reader (Fish 1996 [1973], Attridge 1996 [1987], Weber 1996). But if Jakobsonian methodology is put in the new context of cognitive poetics with its emphasis on the conceptual processing of both the writer and the reader, the status and the rigour of Jakobson’s argument must be reevaluated. At the same time, a Jakobsonian microscopic analysis of the structure of texts will in turn reﬁne the cognitive approach. This chapter is organised as follows: ﬁrstly, how traditional and modern theories have treated metaphor and iconicity is discussed; secondly, the major methodological issues – Peircean semiotics, the theory of blending, and Jakobsonian analysis – are presented and elaborated.

2.1 Background literature 2.1.1 The traditional treatment of metaphor and iconicity Metaphor and iconicity had been treated separately in the traditional literature of linguistics, poetics and rhetoric until the recent development of cognitive linguistics and poetics. Traditional theories of metaphor The most comprehensive review of past studies on theories of metaphor is provided by Ricoeur (1977). He discusses the main viewpoints of metaphorical theories in various disciplines, ranging from linguistics, semantics and the philosophy of language to literary criticism and aesthetics. He classiﬁes such viewpoints into three – rhetorical, semantic and hermeneutic – and identiﬁes his position with the hermeneutic. The rhetorical viewpoint (cf. Aristotle 1954) sees a metaphor as an art of verbal persuasion. A metaphor is regarded as a substitution of words, as completely paraphrasable into a non-metaphorical expression by means of an analogy between the substitution word and the word being substituted. Hence, a metaphor is assumed to be an ornament of language which conveys no new information. Criticising this rhetorical treatment of metaphors, the semantic view takes metaphors to be a transference of meaning. This view can be further divided into the word-centred approach and the

24

Metaphor and Iconicity

sentence-centred one. The former approach is taken, for example, by the semic analysis of Group m (1970). Such word-centred treatment of metaphors contributes to the clariﬁcation of the semantic structure of metaphors. The latter approach is taken by Richards (1936), Beardsley (1958) and Black (1962), who regard a metaphor as an interaction between a metaphorical word and its context. They explicate the mechanism of creating meaning. What is common to both approaches is that they deal with metaphor as a problem concerning meaning in the language system without paying much attention to the inﬂuence of metaphor on extra-linguistic reality. The hermeneutic viewpoint, as taken by Ricoeur (1977, 1978, 1979 [1978], 1981) develops the theory a step further, based on a critical examination of the former two views of metaphor. Ricoeur’s contribution is his emphasis on the power of metaphor to ‘redescribe’ reality. Namely, it is the referential function of metaphorical statements, which breaks through the old categorisation of reality to establish a new one. Thus, his theory foresees the cognitive role of metaphor in conceptualising human experiences and realities. However, his focus is on the creativity of metaphor as an innovation of our view of the world and not on the ubiquity of conventionalised conceptual metaphors, in terms of which we speak, think and act in our ordinary life. Pioneers in the study of linguistic iconicity Only a few linguists since Saussure have discussed the issue of the role of iconicity in the linguistic sign. Jespersen (1933 [1922], 1964 [1921]: 396–411) pioneered the discussion of non-arbitrariness in his study on sound-symbolism. His famous survey on the symbolic value of the vowel [i] shows that this vowel ‘serves very often to indicate what is small, slight, insigniﬁcant or weak’ (Jespersen 1933 [1922]: 283). Works by Fónagy (1963, 1979) present a more comprehensive study in the ﬁeld of sound-symbolism. Bolinger (1965) was interested in word afﬁnities. He claims that there is what he calls ‘sound suggestiveness’ in the sounds shared by a group of words with similar meanings: e.g. the family of ‘slap’, ‘clap’, ‘rap’, ‘tap’, ‘ﬂap’ and ‘lap’ denotes actions that strike and then glide off, or the group of ‘nip’, ‘clip’, ‘tip’, ‘sip’, ‘dip’, ‘grip’, ‘pip’, ‘quip’, ‘yip’ suggests a lighter or sharper blow or its result (cf. 1965: 245–6).1 It was Jakobson who made the most integrated achievements in the treatment of the theme of iconicity before the advent of cognitive linguistics, particularly in his critique of Saussure’s notion of the arbitrariness of the linguistic sign (1971 [1965]). He enlisted Peirce’s typology

Methodological Framework 25

of signs in his argument that language also has the non-arbitrary property of iconicity, i.e. association by similarity between form and meaning in syntax, lexicon, morphology, phonology and poetry.2 His work with Waugh (1979) also devotes a whole chapter to the issue of iconic relations between particular sounds and the meanings they carry. It was also Jakobson who further elaborated the notion of iconicity in the analysis of poetic language, saying that ‘apart from these relatively rare instances of grammatical utilisation, the autonomous iconic value of phonemic oppositions is damped down in purely cognitive messages but becomes particularly apparent in poetic language’ (1971 [1965]: 356) (for further discussion, see Section 2.4). 2.1.2 Metaphor and iconicity in cognitive linguistics and poetics Cognitive linguistics places a new importance on metaphor and iconicity. Both are based on an analogical mapping between different domains. Iconicity deals with a mapping between form and meaning in various degrees of abstractness, whereas metaphor is a mapping between two conceptual domains of meaning, a projection of a schematised pattern from a less abstract source domain onto a more abstract target domain. Although metaphor and iconicity are closely linked cognitive operations (Danesi 1995), they have not been treated as such in the cognitive model of metaphor by Lakoff, Johnson and Turner. As stated above, linguistic discussions of iconicity have been conﬁned to the mimetic relationship between form and meaning. In cognitive semantics, Lakoff and Johnson (1980), Johnson (1987), Lakoff (1987, 1993) and Sweetser (1990) have been more concerned with the centrality of metaphor. Although they have treated iconicity, they have relegated it to a subordinate role. The cognitive theory of metaphor Cognitive linguistics claims that metaphor is primarily an issue of conceptualisation. Metaphors are deﬁned as ‘mappings across conceptual domains’, in which ‘the image-schemata structure of the source domain is projected onto the target domain in a way that is consistent with inherent target domain structure’ (Lakoff 1993: 245). In other words, metaphor allows one to understand a relatively abstract and unstructured subject matter in terms of a more concrete and structured subject matter through image-schemata, which Johnson (1987: 79) deﬁnes as ‘those recurring structures of, or in, our perceptual interactions, bodily experiences and cognitive operations’.

26

Metaphor and Iconicity

Deane (1995: 628–9) provides a succinct summary of the main claims of the cognitive theory of metaphor. Metaphor is deﬁned as a cognitive process in which one set of concepts is understood in terms of another. Each metaphor establishes a mapping between a source domain and a target domain. Metaphoric mappings then cannot alter imageschematic structure of each domain; for example, a part/whole relation in the source domain must correspond to a part/whole relation in the target domain. The mapping thus established enables the transfer of knowledge about the source domain to the target. Many metaphorical mappings are conventionalised in a particular language and culture. The cognitive theory of metaphor assumes that linguistic metaphorical expressions such as ‘I hope our paths will cross sometime soon’, ‘He got a ticket for success’ and ‘they have a bumpy marriage’ are possible because there are metaphorical concepts such as LIFE IS A JOURNEY and MARRIAGE IS A JOURNEY which correspond with the linguistic expressions. In this view, metaphors are primarily a conceptual phenomenon rather than a verbal phenomenon. Metaphors are not classiﬁed as a species parallel to other tropes such as simile, metonymy, synecdoche and so on, but treated as a conceptual base, which gives rise to the linguistic expressions to be classiﬁed as tropes. The cognitive account of poetic metaphor Cognitive poetics3 (cf. Turner 1987, 1996, Lakoff and Turner 1989, Freeman 2000, among others) applies the cognitive theory of metaphor as outlined above to the analysis of poetic texts. It claims that the basic conceptual metaphors, which underlie everyday expressions, also underlie many of the poetic metaphors and that they serve in part to give power to the poetic metaphor. In other words, the poetic metaphor is a non-conventional use of conventional metaphor. Lakoff and Turner (1989: 67–72) explicate the ways in which poets aim at poetic effects from conventional cognitive metaphors. According to them, there are four techniques: extending a conventional metaphor in a novel way; elaborating the image-schemas by ﬁlling special or unusual cases; questioning the limitations of conventional metaphors and offering a new one; and forming composite metaphors by the non-conventional combination of multiple conventional metaphors for a given target domain. These techniques, which use conventional metaphors in special and non-automatic ways, make poetic metaphor noticeable and memorable. Poets, by their non-conventional use of metaphors, lead their readers beyond the bounds of ordinary modes of thought.

Methodological Framework 27

Several studies have contributed to revealing how the poetic text represents the conceptual universe of such metaphors. As outlined above, Lakoff and Turner (1989) have pioneered our understanding of the way in which cognitive metaphors structure poetic texts, and the way in which poetic texts are structured as a metaphorical icon of the overall meaning that they represent. Research by Turner (1987, 1996) has demonstrated in detail how literary metaphors use the ordinary system of conventional conceptual mappings. Analyses of English poems from a cognitive perspective include Deane (1995) on W. B. Yeats, M. Freeman (1995, 1997, 2000) on Emily Dickinson, Hamilton (2003) on Wilfred Owen and Steen (2003) on love poems. Works by D. Freeman (1993, 1995) offer a comprehensive analysis of conceptual metaphors in Shakespearean drama, and Sweetser (1995) provides a cognitive metaphorical account of Greek mythology. Gavins and Steen (2003) and Semino and Culpeper (2002) are collections of papers on the cognitive analysis of literary texts ranging from poetry, parable and short stories to narrative texts. Cognitive studies of poetry have clariﬁed the fact that poetic texts can be metaphoric on two levels. On the one hand, they display local metaphors, which manifest themselves locally as linguistic expressions in the text. They are based on either conceptual mappings (systematic mappings at the conceptual level such as events as actions, and life as journey), image mappings (mapping of the motor-sensory images evoked by linguistic expressions in the source superimposed onto the target, e.g. the colour of objects mapped onto feelings), or a combination of both. On the other hand, texts as a whole can be read metaphorically. This is the case of text as global metaphor. Namely, the text presents a source which is to be mapped onto a target of larger concern. According to Lakoff and Turner (1989: 146–7), a reading of the text as a global metaphor is constrained in three major ways: by the use of conventional conceptual mapping; by the use of commonplace knowledge; and by iconicity – a mapping between the structure of a text and the meaning or image it conveys. Cognitive poetics therefore sees iconicity as a structural mapping between the form and meaning of the text, and metaphor as a system mapping between the conceptual domains mediated by the text. The cognitive treatment of iconicity The problem of iconicity is indeed undergoing a revival in cognitive studies of syntax, morphology, lexicon, language change and pragmatics. Here the iconic motivation of form with respect to meaning has

28

Metaphor and Iconicity

been investigated with reference to the correspondence between conceptual structure and linguistic manifestations in grammar and discourse. Haiman (1985a) offers a detailed survey of iconicity in syntax, while works by Waugh (1992, 1994, Waugh and Newﬁeld 1995) present a study on iconicity in morphology and lexicon. There have been several collections of papers on iconicity published since 1985. Haiman (1985b) is one of the ﬁrst collections in contemporary linguistics, including articles on diagrammatic iconicity in various linguistic phenomena, such as iconicity in inﬂectional forms, temporal sequence, grammatical categories, intonation and deixis. Hiraga and RadwanskaWilliams (1994) is a special issue featuring iconicity and metaphor in syntax, morphology and pragmatics. Landsberg (1995) collects more recent works on iconicity in syntax. This growing interest in the issue of iconicity in cognitive linguistics is having repercussions in the related ﬁelds of semiotics and anthropology. Bouissac et al. (1986), Evans and Helbo (1986) and Simone (1995) have articles on various aspects of iconicity from a semiotic perspective. Hinton et al. (1994) is a collection of papers on soundsymbolism, representing a recent accumulation of cross-linguistic data. Nänny and Fischer (1999) and Müller and Fischer (2002) collect articles on iconicity from a range of different perspectives including semiotics, linguistics, graphics and literature. Iconicity has of course found its detractors. Among the most prominent of those detractors is the semiotician, Umberto Eco. Eco (1976, 1984) emphasises the importance of cultural codiﬁcation for any signfunction. He sees that even in iconic signs a conventionally established code is in operation. Rejecting the notions of iconicity such as a sign sharing the same properties as its object, being similar or analogous to its object, or being motivated by its object, Eco (1976: 192) assumes that icons are ‘culturally coded without necessarily implying that they are arbitrarily correlated to their content’ (italics in the original). However, Eco’s argument could be appropriated by the Peircean analysis of iconicity elaborated below, in which it is stated that diagrammatic and metaphorical icons may involve convention (for further discussion, see Section 2.2.3).

Having in mind the background literature on metaphor and iconicity, I will move to major methodological issues, which give a frame of reference to the present study. The deﬁnition of metaphor and icon by Peirce will be reformulated with the aid of the contemporary development of the cognitive theory of metaphor.

Methodological Framework 29

2.2 Peircean deﬁnition The semiotic notion of iconicity originates with the American philosopher Charles Sanders Peirce. In order to grasp the Peircean notions of iconicity and metaphor, it is essential to examine them in the entire context of his doctrine of the sign, which is beyond the scope of this study. However, Peirce’s persistent classiﬁcation of the signs into three categories that he calls Firstness, Secondness and Thirdness deserves a brief summary as it is this ternary account that distinctively marks Peircean semiotics and recurs throughout his work (cf. Feibleman 1946, Ayer 1968). It is also this aspect of Peirce’s system that is particularly relevant to the treatment of metaphor and iconicity in this work. My claim that metaphor and iconicity are interrelated is derived from Peirce’s ternary account whereby the notion of Thirdness includes both Secondness and Firstness, the notion of Secondness includes Firstness, while the notion of Firstness stands alone. It is the idea of Thirdness as an inclusive category of the other two that is of most importance in my analysis of metaphor–icon links. 2.2.1 Firstness, Secondness, Thirdness Peirce deﬁned the sign as follows: ‘a sign, or representamen, is something which stands to somebody for something in some respect or capacity’ (italics in the original) [2.228].4 That is, a sign A stands for an object B to an interpretant C. This simple formulation of the deﬁnition of the sign characterises the ternary structure and the interactive nature of the signs that Peirce sought out in his semiotic theory. A sign, or representamen, is ‘the First Correlate of a triadic relation, the Second Correlate being termed its Object, and the possible Third Correlate being termed its Interpretant’ (italics in the original) [2.242]. There are two critical points to note: the basic traits of these three modes; and the importance of the mediatory nature of the third. The distinctions of Firstness, Secondness and Thirdness can be drawn in terms of their basic traits. For Peirce [1.23], these distinctions represent three different ontological modes of being, deﬁned in his words as ‘possible quality’, ‘actual fact’ and ‘law in the future’. Elsewhere [6.32], he deﬁned them according to the relations of elements they involve. Firstness is the monadic mode, in which signs can exist independently of anything else; Secondness presupposes the dyad, in which two elements are compared or connected, such as cause and effect. Thirdness is the ‘conception of mediation whereby a ﬁrst and a second are brought into relation’. As regards the relationship among these three modes,

30

Metaphor and Iconicity

Peirce said, ‘not only does Thirdness suppose and involve the ideas of Secondness and Firstness, but never will it be possible to ﬁnd any Secondness or Firstness in a phenomenon that is not accompanied by Thirdness’ [5.90]. It is clear that Peirce regards Thirdness as embracing both Firstness and Secondness. In my view, this mediatory nature of Thirdness is particularly important in the application of the Peircean classiﬁcation to the analysis of sign systems, because Thirdness functions as a general law, which navigates the interpretation of the Firstness of ‘quality’ and the Secondness of ‘actuality’. In other words, Thirdness provides a context or a relation in which the other modes of being can be identiﬁed and interpreted. 2.2.2 Icon, index, symbol It is by the application of the three categories above to the sign–object relation that the icon–index–symbol distinction emerges in the Peircean theory of signs, and so does the image–diagram–metaphor distinction. Peirce deﬁned signs in relation to their objects according to his Firstness, Secondness and Thirdness and termed them icons, indices and symbols. Table 2.1 indicates the deﬁning conditions and the means by which these three signs achieve their signiﬁcation. An icon is deﬁned as a sign which represents an object mainly by its similarity to that object. An icon is a Firstness of the three signs, because its signiﬁcation is achieved by its intrinsic characteristics. For example, a photograph is an icon of a person photographed when one sees the similarity between them or when it is used as such. ‘Anything whatever, be it quality, existent individual, or law, is an Icon of anything, in so far as it is like that thing and used as a sign of it’ [2.247]. A sign of Secondness, an index, represents its object by its existential relation to the object. In Peirce’s words, ‘an Index is a sign which refers to the Object that it denotes by virtue of being really affected by that object’ (italics in the original) [2.248]. There is a contiguity relation, either spatial or causal, between the sign and the object. For example,

Table 2.1 Icon, index and symbol Sign

Deﬁning conditions

Signiﬁcation by virtue of

Icon Index Symbol

Similarity Contiguity (causal or spatial) Conventionality

Its intrinsic characteristics Its existential relation to the object Its relation to the interpretant

Methodological Framework 31

smoke is an index of ﬁre, based on the causal relation in reality. Smoke exists as an effect of ﬁre, and therefore it indexes ﬁre. ‘A Symbol is a sign which refers to the Object that it denotes by virtue of a law, usually an association of general ideas, which operates to cause the Symbol to be interpreted as referring to that Object’ (italics in the original) [2.249]. For example, a red trafﬁc light is a symbol of stopping for those who know the trafﬁc rules. As Peirce mentioned, any utterance of speech ‘signiﬁes what it does only by virtue of its being understood to have that signiﬁcation’ [2.304]. It is undeniable that language structure and use depend largely on symbolic signiﬁcation. As one cannot play a game of chess when one does not understand its rules, one cannot speak a language when one does not understand its rules. In dealing with the examples, however, what seems important to me is the nature of manifestations of these signs rather than the nature of classiﬁcation as an ideal and abstract construct. Even when it is granted that a photograph is iconic, and that language is symbolic, it can still be argued that a photograph can be an index because a person photographed is a cause and the photograph is an effect, or that there are indexical (e.g. pronouns, demonstratives) and iconic (e.g. soundsymbolism, word order) aspects in language (cf. Jespersen 1964 [1921], Jakobson 1971 [1965], Haiman 1985a, b). In the real sign–object situations, a sign may involve all three types of Peircean signs – icon, index and symbol – with predominance of one over the others. Moreover, when one looks for examples to explain any of his signs to the full, it is clear that both Firstness and Secondness can be involved in Thirdness. Peirce said, ‘the category of ﬁrst can be prescinded from second and third, and second can be prescinded from third. But second cannot be prescinded from ﬁrst, nor third from second’ [1.353].5 2.2.3 Image, diagram, metaphor It is precisely in this context of Peircean ontology that his further divisions of icons are to be introduced, i.e. the divisions of icons into images, diagrams and metaphors. Again, the trichotomy rules the classiﬁcation, and subtypes of icons are characterised by the dominance of mimicry, analogy and parallelism. Those which partake of simple qualities, or First Firstness, are images; those which represent the relations, mainly dyadic, or so regarded, of the parts of one thing by analogous relations in their own parts, are diagrams; those which represent the representative character of a

32

Metaphor and Iconicity

representamen by representing a parallelism in something else, are metaphors. (Italics in the original) [2.277] Image achieves similarity by partaking of some of the simple qualities of its object, such as a portrait for the person portrayed or a sample of a wallpaper for the wall. It is First Firstness, because it is an intrinsic aspect (First) of the intrinsic quality (Firstness) that characterises images. The relationship between an image and its object is monadic in the sense that the deﬁning condition of similarity is achieved by the image’s exhibiting its intrinsic qualities which resemble some of the simple qualities of its object. In short, images are mimetic to their objects. Images are immediate signs, which are less abstract and less dependent on conventions than the other two subtypes of icons. The second subtype is the diagrams, which exhibit a structure analogous to the structure of their object. Diagrams, as Second Firstness, show the abstract structure which resembles the structure of their object. Diagrams are, in this sense, analogical to their objects in structure, but not necessarily in substance. For example, an architectural elevation shares no signiﬁcant substantial properties with the actual building; nevertheless, it may, in conjunction with certain conventions, be graphically representative of the structural relationships among the elevational designs of the building. Or a scientiﬁc model of the solar system is a diagram of the actual solar system. The model is proportionally much smaller than the actual object and the materials composed of the model are different from the actual solar system; but it represents analogically the relational positions of planets in the actual solar system. As the diagram involves an analogical relationship between the two abstract structures, i.e. that of a sign and of an object, the relationship between the diagram and its object is dyadic. The diagrammatic signiﬁcation is a dyadic correspondence or isomorphism, which is characterised by the structural analogy between the diagram and its object. Moreover, the similarity between the two structures is univocal rather than equivocal. Although it is on an abstract level, there should be oneto-one correspondence between the dyads, which is the basis of the structural analogy. In a word, the similarity between the diagram and its object is a dyadic, relational or structural or proportional analogy. Being based on a structural rather than a sensory dimension, diagrams are more abstract and conventionalised than images. It might be said that the image is recognisable by anyone, whereas to recognise the diagram requires an understanding of certain conventions. The third subtype of iconicity, metaphor, is only very brieﬂy discussed

Methodological Framework 33

by Peirce. Besides the deﬁnition given above, there are only a few descriptions of metaphors in his writings, in contrast to his recurrent explanation and description of images and diagrams. Metaphors, as I interpret them from the deﬁnition, ‘those which represent the representative character of a representamen by representing a parallelism in something else’, are different from images and diagrams in that they require an existence of ‘something else’, i.e. a third thing in addition to a sign and the object. In this sense, it is by a triadic relation that metaphors achieve their signiﬁcation. For example, a metaphorical icon (e.g. ‘The human mind is a computer’) signiﬁes its object (e.g. ‘mind’) by pointing to a parallelism between the object (e.g. ‘mind’) and something else (e.g. ‘computer’). This parallelism could be a set of semantic associations (e.g. receiving ideas, storing ideas, manipulating ideas) between the object and something else. Instead of saying that ‘there are too many ideas in my mind’, the metaphorical message, ‘I have too many ﬁles open in my mind’, points to an associative parallelism in which a set of characteristics of a computer is mapped onto that of the human mind. Because of its triadic mediatory nature, i.e. Thirdness, a metaphorical sign largely depends on convention. This is particularly so when metaphors become conventionalised, as they manifest themselves in everyday language (cf. Lakoff and Johnson 1980). On the other hand, a novel and creative metaphor requires both imagination and convention in order for us to make sense out of it (cf. Lakoff and Turner 1989). For such a metaphor manifests both the feeling and the law, i.e. Firstness and Thirdness. Furthermore, when a metaphorical text displays itself as an allegory, then it involves an indexical interpretation, pointing to an actual context it relates to, i.e. Secondness. It is speculated, therefore, that a metaphor is a sign which exhibits Firstness (quality), Secondness (fact) and Thirdness (law). Following Peirce’s logic of Thirdness, which, in his characterisation, includes Firstness and Secondness (cf. [5.90]), it can be said that metaphor also involves image and diagram, and so does symbol involve icon and index. As Haley (1988: 34) convincingly puts it, the relations among the three hypoicons may be more aptly conceived as a continuum of iconicity, suggesting that in every iconic experience there is at least some degree of metaphoricity, diagrammaticality, and imagery; the three grounds of iconicity would then be distinguished, in actual occurrence, according to which ground was judged as predominant. (Italics in the original)

34

Metaphor and Iconicity

Table 2.2 Subtypes of icon6 Subtypes

Image

Diagram

Metaphor

How similarity is achieved

Partakes of some of the simple qualities of its object

Exhibits the abstract structure of its object

Represents a parallelism in something else

Similarity in

Quality

Structure

Association

Sign–object (sign) relationship

Monadic immediate mimicry

Dyadic structural or relational analogy

Triadic representational parallelism

Example

Sample

Model

Parable

Manifestations in language

Onomatopoeia, sound-symbolism

Linearity, proximity, symmetry, asymmetry, etc.

Metaphorical transfer from one domain to another

The immediacy of sign–object link decreases from images to diagrams, from diagrams to metaphors. At the same time, an icon may involve all the three subtypes with predominance of one over the others. The differences among the three subtypes of icons are a degree of abstraction and conventionality as well as dominance of different characteristics of iconicity, as illustrated in Table 2.2. To sum up, images are based on a monadic simple, sensory or mimetic resemblance; diagrams are based on a dyadic proportional or structural analogy; and metaphors are based on a triadic mediatory representational parallelism. The degree of abstraction increases from images to diagrams, from diagrams to metaphors. Although Peirce emphasised the importance of the third mediatory category, he never developed his notion of metaphor in relation to images and diagrams. As brieﬂy mentioned in Chapter 1, this lack of elaboration might be derived from the notional confusions in the three subtypes. For the ﬁrst two, images and diagrams, deal with a similarity relation between form and concept, while metaphors deal with a similarity relation between two concepts. In the following section, I am redressing Peirce’s neglect of metaphor in his ternary account of iconicity, by linking his understanding to the contemporary treatment of metaphor. Peirce’s insight is therefore reﬁned with the aid of the cognitive theory of metaphor.

Methodological Framework 35

2.3 Cognitive deﬁnition: interplay between metaphor and iconicity 2.3.1 Iconicity and metaphor deﬁned The cognitive approach regards metaphor and iconicity as important theoretical issues in the exploration of the operation of the human mind. The vagueness of the Peircean notion of metaphor is compensated for by the deﬁnition given by the cognitive theory of metaphor. Working deﬁnitions In cognitive terms, iconicity (i.e. images and diagrams in Peircean terminology) refers to the mapping between form (structure) and meaning (cf. Lakoff and Turner 1989) in various degrees of abstraction, from concrete attributive resemblance to abstract relational analogy. Metaphor, on the other hand, is the mapping between two conceptual spaces of meaning, a projection of a schematised pattern (system) from a less abstract source space onto a more abstract target space (cf. Lakoff and Johnson 1980, Lakoff 1987, 1993, Turner 1996). In most cases, the basis of similarity between icons is derived from visual, auditory and other formal traits of the object that they stand for, while iconic signs are themselves often visual, auditory or formal representations. Metaphors, however, do not necessarily have such traits, but act as a heuristic device to mediate dissimilar concepts by means of the similarity they yield. As Anderson (1984: 459) correctly points out, ‘a metaphor, like an image or an analogy, is what it represents – but not because of an antecedent identity or similarity, not as a reminiscence, but in virtue of a similarity which it creates’. Indeed, in icons, the similarity relationship between the sign and the object is taken to be preexistent. For example, there is a similarity between the folder icon on a computer screen and the folder object in the ofﬁce even if one has no prior experience with computers. In contrast, a metaphor connects two entities, say a mind and a computer, which are a priori dissimilar. The connecting act via a metaphor establishes the similarity between two dissimilar entities, and thus creates a new meaning or interpretation. Metaphor and iconicity often work together, through what can be called an iconic moment in metaphor. Such iconic moments manifest themselves as image-schematic structures that generate speciﬁc meaning. In addition, there needs to be some metaphorical navigation in the understanding of icons. Hence, metaphor in icons concerns an interpretation of iconic aspects of the form. The metaphors that

36

Metaphor and Iconicity

underlie the relationship of form and meaning (i.e. grammatical metaphors, see Section 2.3.3 below) guide the interpretation of forms as diagrammatic icons. This interplay between metaphor and iconicity is not a static product of structure, but a dynamic process. It emerges from, is elaborated by and integrated with background knowledge, contextual information and human feelings. This takes place at the precise time at which meaning is generated both in composition and interpretation. 2.3.2 The model of blending Although cognitive linguistics offers a clearer deﬁnition for metaphor and iconicity, it is not satisfactory that metaphor and iconicity are treated as somewhat isolated manifestations. Because metaphor and iconicity are similar cognitive operations motivated by analogical reasoning, they would be best captured in their interplay rather than in isolation as Peirce correctly suggested but failed to develop. Therefore, my intention is to refocus Peirce’s original conceptualisation of metaphor and other iconic signs in the context of the cognitive theory of blending. The claim is that the model of blending developed by Turner and Fauconnier (Turner 1996, 1998, Turner and Fauconnier 1995a, b, 1999, Fauconnier and Turner 1998, 2002, among others) offers a promising theoretical basis and an effective methodological contribution to explicating the interrelationship of metaphor and iconicity (i.e. Peircean ‘image’ and ‘diagram’) and analysing their actual manifestations in language. Basic claims of the model of blending Turner and Fauconnier (1995a: 184) propose a model of ‘conceptual projection across four or more (many) mental spaces rather than two domains’, to explain a wide range of phenomena including ‘conceptual metaphor, metonymy, counterfactuals, conceptual change’ (p. 183), ‘classiﬁcation, the making of hypotheses, inference, and the origin and combining of grammatical constructions’ (p. 186), ‘idioms, . . . , jokes, advertising, and other aspects of linguistic and nonlinguistic behavior’ (p. 186). The basic claims of the cognitive account of metaphor were expounded by Lakoff, Johnson and Turner (Lakoff and Johnson 1980, Lakoff 1987, 1993, Lakoff and Turner 1989), and elaborated as a more general model of ‘blending’ by Turner and Fauconnier (Turner and Fauconnier 1995a, b, Turner 1996, Fauconnier and Turner 1998, 2002, among others). These ideas are summarised as follows and illustrated in Figure 2.1:

Methodological Framework 37 Generic Space

Input (Source)

Input (Target)

Blend Figure 2.1 The model of blending7

1. Metaphor is a cognitive process in which one set of concepts (a target) is understood in terms of another (a source). According to the model of blending, metaphor is a conceptual integration of four mental spaces.8 Mental spaces are small conceptual arrays constructed for local purposes of understanding. When a conceptual projection occurs, two input mental spaces (source and target in a metaphor) are created. There is a partial mapping of counterparts between the input spaces, as represented by solid lines in Figure 2.1. These input spaces have relevant information from the respective domains, as well as additional structure from culture, context, point of view and other background information. 2. There are two kinds of middle mental space in addition to the input spaces. These middle spaces are a generic space and a blended space. The generic space maps onto each of the input spaces, as indicated by dashed lines in Figure 2.1. It reﬂects some abstract structure shared by the inputs. The blended space also receives partial projections from the inputs, as indicated by dashed lines. The blend is a rich

38

Metaphor and Iconicity

space integrating the generic structure, structures from each input space and background information. Often the blended space has an emergent structure of its own, as represented by the square in the ﬁgure. This emergence occurs in ways of composition (in which new relations emerge from projections from the inputs), completion (in which the composite structure projected into the blend is completed into the larger system by background knowledge, cognitive and cultural models), and elaboration (in which the blend is further elaborated according to its own logic) (cf. Fauconnier and Turner 1998: 271). 3. Each mental space has a consistent image-schematic structure that is preserved through a conceptual projection of generic and input spaces. The image-schemas are skeletal patterns in our sensory and motor experience, such as a container, a motion along a path, part and whole, centre and periphery, symmetry and so forth. 4. The blended space develops inferences, arguments, ideas and emotions, which can modify the initial input spaces and change our views of the knowledge used to build those input spaces (cf. Turner 1996: 83). Therefore, the blend is a dynamic mechanism of creativity. Iconic moments in metaphor By using the example, ‘I have too many ﬁles open in my mind’, I will recapitulate the main points about iconic moments in metaphor according to this model of blending. As discussed above, the conceptual metaphor, MIND IS A COMPUTER, allows us to understand an abstract entity, a human mind, in terms of a concrete entity, a computer. The cross-space mapping between the inputs constitutes the content of the generic space. In the above example, the image-schemas of objects resembling ﬁles, an excessive amount of them, a machine and its movement in the source input (‘too many opened ﬁles’) are aligned with the imageschemas of abstract objects and manipulation of the mass of these objects in the target input (‘too many ideas’) by the generic space which has the highly abstract, skeletal image-schematic structure taken to apply to both the input spaces (e.g. a frame of an excess of objects). It should be noted that there are iconic moments in this alignment of image-schemas. The choice of the term, ‘image-schema’, itself suggests that both ‘image’ and ‘diagram’ are to be related to this cross-space mapping of metaphor. In the input spaces, from a vast amount of information provided by our bodily experiences, interactive perception, contextual and background knowledge, we have an array of images that do

Methodological Framework 39

not clash with the entire cognitive process of metaphor. In our example, the image content includes visual (or interactive) images of an action of opening many ﬁles, a screen showing many ﬁles open, a dialogue box saying ‘too many ﬁles open’, and/or a screen frozen due to too many ﬁles open. These are mimetic mental representations of sensory perceptions, and constitute imagic iconicity. At the same time, a mental space develops a structure by selecting and schematising the images, namely, an image-schematic structure, which has a diagrammatic representation of the image content of mental space (e.g. an image-scheme of a blocked action due to an excessive amount of ﬁles). A similar process might occur in the target input space to a less concrete degree. For example, the evoked images for ‘too many ideas’ may be more abstract than what happens in the source, and so is its schematisation. Therefore, there are two types of iconic mappings in the input spaces: an attributive imagic mapping between the sensory perception and the image content, and an analogical diagrammatic mapping between the image content and the schematised images. Conceptual and image projections not only occur in input spaces, but they also include a generic space and a blended space. There is a relationship of similarity between the generic space and the input spaces, because the generic space has an image-schematic structure shared with the two inputs. This shared image-schematic structure is skeletal and abstract. In the metaphor, ‘I have too many ﬁles open in my mind’, the generic structure can be an image-schema of a purposive action or some kind of blocked progress. As the image-schematic structure is preserved, the two input spaces are structurally analogous. The analogy supported by the generic space can be described as diagrammatic because it is a structure-preserving schematic correspondence between the source and the generic space, and between the target and the generic space. In other words, diagrammatic iconicity is at work in the analogy between the corresponding image-schematic structures of the generic space and the input spaces. This metaphor gets its full interpretation and understanding as a conceptual integration in the blended space, which has a conceptual mapping of partial structures from both inputs and the generic space, and which develops an emergent structure of its own, for example, disorder, a lack of proper motion, a stoppage, as a result of excessive movements. Figure 2.2 illustrates imagic and diagrammatic iconic mapping inherent in the conceptual mapping and integration of metaphor. Needless to say, it is metaphor rather than iconicity that prevails in this particular example of ‘I have too many ﬁles open in my mind’. For

40

Metaphor and Iconicity

Generic Space DIAGRAMMATIC MAPPING

DIAGRAMMATIC MAPPING

Image-Schema

Target

Source DIAGRAMMATIC MAPPING Image Content IMAGIC MAPPING Sensory Perception

Blend CONCEPTUAL MAPPING AND INTEGRATION

METAPHOR Figure 2.2 Metaphor–icon links in the model of blending

only the very abstract conceptual (theory-building) level seems to involve iconic moments: imagic and diagrammatic iconicity in the evoking of images and the schematisation of images in the input mental spaces, and diagrammatic iconicity in the correspondence or counterpart projection of image-schematic structures between the generic space and input spaces. The discussion above has illustrated that the model of blending lends itself to understanding the semiotic implications of imagic and diagrammatic iconicity in the cognitive process of metaphor. When I say that metaphor has iconic moments, it means that there are some traces of iconic mapping in general between a fairly abstract form and a

Methodological Framework 41

concept. No visible ﬁgure mirrors the concept in the linguistic representation of this particular example of a conventional metaphor. This does not mean that language structure and use are deprived of iconicity. Until linguistic iconicity is delineated, however, it will not become visible, because it is not foregrounded. 2.3.3 Grammatical metaphors Iconic meanings given to form by metaphor One of the ways by which we can detect iconicity in linguistic resources (such as sounds, words, sentences, discourses and writing systems) comes from conventional metaphors relating to our conception of language. Lakoff and Johnson (1980: 126–38) provide a basic insight into how metaphor can give meaning to form. They point out that we conceptualise language by orientational and ontological metaphors, that is, in terms of space and in terms of objects. A pervasive conceptualisation of language9 is the CONDUIT metaphor (Reddy 1979) in which ideas are seen as objects, linguistic expressions as containers, and communication as sending and receiving the containers containing ideas as objects. Because we speak and write in a linear order, we also conceptualise language in terms of space in a linear order. There are a handful of conventional conceptual metaphors which function as a useful tool to give iconic meanings to form. Table 2.3 gives a non-exhaustive list of examples of such conventional metaphors. For ease of discussion, I will call these conventional metaphors ‘grammatical metaphors’, because these metaphors, more or less, concern the relationship of form and meaning in grammatical conventions (e.g. phonology, morphology, word formation and word order). Because we conceptualise expressions as containers and meanings as contents, it is assumed that the bigger containers have the larger contents. So, MORE MEANING IS MORE FORM. For example, prolongation, repetition, iteration and reduplication of linguistic forms such as a sound, a syllable and a word tend to stand for more of the content. The SIMILARITY OF FORM stands for the SIMILARITY OF CONTENT, the DIFFERENCE OF FORM, the DIFFERENCE OF CONTENT. For example, the use of the similar sound signiﬁes the similar (or equivalent) meaning in English morphology, an [s] for the plural, a [non] for the negative preﬁx and so forth. LOCATION OF FORM also gives iconic meanings. In general orientational experiences, we put important things in the centre. For example, there tends to be a building or a monument of importance (a castle, a

42

Metaphor and Iconicity

Table 2.3 Grammatical metaphors Source

Target

QUANTITY OF FORM

QUANTITY OF CONTENT MATERIAL

SIMILARITY OR DIFFERENCE OF

LOCATION

SIMILARITY OR DIFFERENCE OF CONTENT

SEQUENCE SYMMETRY/ASYMMETRY OF FORM

SYMMETRY/ASYMMETRY OF CONTENT

LOCATION OF FORM

CLOSE

STRENGTH

LINEAR

SEQUENCE OF TIME

LINEAR

CAUSE/EFFECT

FIRST/LAST

IMPORTANCE

SEQUENCE OF FORM

church or a tower) at the centre of a city, a village or a campus. In English, for instance, when a word is embellished with inﬂections and afﬁxes, they are peripheral. The stem, which expresses the core meaning, stays in the central position. It is rare that the central shape of a word is altered. Not only the location of the form, but also the sequential order of the form can mirror the space, time and causal sequence of the content. In the standard representation of temporal events, for example, it is more common to state things in line with their occurrence in time, from past to the present rather than the reverse. Halliday (1994b) has also noted the diagrammatic correlation (which he calls ‘iconic’) between expressions in the discourse grammar and the contents expressed. According to him, there are three types of such correlation: 1. There is a movement from a given Theme (background) to a rhematic New (foreground); this movement in time construes iconically the ﬂow of information.

Methodological Framework 43

2. New semiotic entities are created by . . . nominal packages, . . . ; the nominal expression in the grammar construes iconically an objectiﬁed entity in the real world. 3. The combination of (1) and (2) construes iconically the total reality, . . . a reality consisting of semiotic entities in a periodic ﬂow of information (1994b: 143). Halliday’s notion of ‘packages’ and ‘movement’ is similar to the grammatical metaphors discussed above, namely, LINGUISTIC EXPRESSIONS ARE CONTAINERS and CHRONOLOGICAL OR CAUSAL ORDER IS A LINEAR SEQUENCE; but it goes further to point out that nominal expressions are the prototypical construal of EXPRESSIONS as PACKAGES. Structural diagram and relational diagram As a diagram refers to its object by virtue of similarity between the relationship among the parts of the diagram and the relationship among the parts of the object, the structure of linguistic representation resembles the structure of the content that it conveys. As in Caesar’s ‘veni, vidi, vici’ (I came, I saw, I conquered) (cf. Jakobson 1971[1965]: 350), the syntagmatic order of mention in speech corresponds to the chronological or causal order in which the event occurs. Paradigmatically, word afﬁnity relations such as morphemes (e.g. ‘acceptable’, ‘readable’ and ‘replaceable’ – sharing ‘able’ in form and ‘being CAPABLE of doing’ in meaning), phonesthemes (e.g. ‘gleam’, ‘glance’, ‘glare’ and ‘glitter’, – sharing initial /gl/ in form and ‘connection to vision’ in meaning) and so forth, suggest that similarity in form signals similarity in meaning; difference in form signals difference in meaning. Diagrammatic icons are, in this sense, analogous to their objects in structure and/or in relation. Semiotically speaking, these two examples illustrate two broad possibilities in which diagrammatic iconicity is manifested in grammar, and thereby receives interpretation as such by the presence of ontological and orientational metaphors explained above. I will call the two types of iconic form ‘structural diagram’ and ‘relational diagram’, which roughly follow Haiman’s classiﬁcation of diagrammatic iconicity into ‘motivation’ and ‘isomorphism’ (cf. Haiman 1985a, b).10 ‘Structural diagrams’ display a correspondence between structure of form and structure of content, whereas ‘relational diagrams’ show a tendency to associate similarity in form with similarity in content; difference in form with difference in content. In an extreme case, this tendency is expressed as the principle of ‘one meaning, one form’ (cf. Bolinger

44

Metaphor and Iconicity

Iconic Mapping

SEQUENCE IN TIME IS SEQUENCE IN FORM

QUANTITY OF CONTENT IS QUANTITY OF FORM

FORM

MEANING

Figure 2.3 Structural diagram

1977). The difference between structural diagrams and relational diagrams seems to be that the former are a structural analogy whereas the latter are a relational analogy. Structural diagrams tend to deal with a correspondence between the structure of linguistic form and the structure of conceptualisation; relational diagrams with a correspondence between the relation in linguistic form and the relation in linguistic meaning. Figures 2.3 and 2.4 are a graphic presentation of the two principles in question. Both auditory and visual forms can receive an iconic meaning from the grammatical metaphors, and function as structural or relational diagrams. The relationship of form and meaning in grammatical metaphorical mapping is diagrammatic, because what is preserved is an analogical relationship mediated indirectly by grammatical metaphors. This contrasts with a direct attributive connection such as pure imagic iconicity between the linguistic form and meaning, e.g. a case of onomatopoeia

Methodological Framework 45

Iconic Mapping

DIFFERENCE IN MEANING IS DIFFERENCE IN FORM

SIMILARITY IN MEANING IS SIMILARITY IN FORM

FORM

MEANING

Figure 2.4 Relational diagram

and visual language such as logographs (for more examples and discussion on grammatical metaphors and diagrammatic manifestations, see Chapter 6).11

2.4 Jakobson on poetic function revisited In addition to the theoretical basis on which I investigate manifestations of metaphor–icon links, I would like to provide some methodological considerations of how to analyse the texts. As discussed in Chapter 1, the primary data of analysis for metaphor–icon manifestations will be sought in Japanese and English poetic texts. Poetic texts

46

Metaphor and Iconicity

are considered to offer optimal examples to explore the issue and clarify the complexity of metaphor–icon links because their manifestations are foregrounded as the ‘poetic function’ of language. 2.4.1 The poetic function of language In twentieth-century ‘linguistic poetics’ (cf. Fónagy 1965, Jakobson 1971, 1981, 1985 [1980], Shapiro 1976, Worth 1977, Jakobson and Waugh 1979, Jakobson and Rudy 1985 [1980], among others), the structure of poetry has been analysed by looking at certain abstract patterns in the distribution of linguistic elements, such as distinctive features, phonemes, syllables, morphemes, words, phrases and sentences. The subliminal patterns revealed by these microscopic analyses have supported Jakobson’s assertion that ‘the poetic function projects the principle of equivalence from the axis of selection onto the axis of combination’ (Jakobson 1960: 358). When two elements in the sequential conﬁguration are equated due to some kind of formal similarity such as alliterated words, rhymed feet or syntactic parallelism, and when there is also a semantic or conceptual link between the two elements thus equated, then it can be said that they exhibit an iconic manifestation: similarity in form signals similarity in meaning. Thus, what Jakobson refers to as the principle of equivalence can be interpreted as a way to analyse the manifestations of iconicity as he claims that ‘in poetry similarity is superimposed on contiguity, and hence “equivalence is promoted to the constitutive device of the sequence”’ (Jakobson 1985[1960]: 42). 2.4.2 The role of repetition in text The most prominent patterns revealed by the semiotic analysis of poems tend to take some form of the recurrence of equivalent elements, namely, a repetition of some kind. It seems inevitable that an analytic methodology of structural poetics leads to a realisation of such recurrent patterns, because the working principle of equivalence, when projected onto the axis of combination (Jakobson 1960: 358), presupposes the recurrence of equalised forms. Traditionally, repetition in poetry has been studied as a rhetorical device in such forms as alliteration, assonance, consonance, rhyme, meter and so on. Attention has been paid to phonological and prosodic aspects with less emphasis on the semantic. The Jakobsonian approach, however, tackles the semantic effects of repetition, namely, the semantics of patterns of formal parallelism revealed by a detailed semiotic

Methodological Framework 47 Table 2.4 Units of repetition Visual level

Auditory level

Conceptual level

A part of character Character Part of lexical item Lexical item Phrase Sentence Grammatical pattern

Distinctive feature Phoneme, syllable, mora Syllable, mora, ictus Lexical item Phrase Sentence Grammatical pattern

Radical Logograph Morpheme Lexical item Phrase Sentence Grammatical pattern

analysis of a poetic text. Concrete and abstract repetitive patterns are manifest in several general types, and carry different types of iconic meaning. Units of repetition Units of repetition can be detected from linguistic resources. They vary in length and appear at different levels such as visual, auditory and conceptual levels, as indicated in Table 2.4. The visual level has a part of character, a character, a part of a lexical item, a lexical item, a phrase, a sentence and a grammatical pattern. The auditory level, corresponding to the visual units of ‘part of a character’ up to ‘part of lexical item’, has a distinctive feature, a phoneme, syllable, mora, or ictus. On the conceptual level, corresponding to these three smallest units are a radical, a logograph and a morpheme. Once one gets to the level of a complete lexical item, the visual, the auditory and the conceptual levels have a one-to-one correspondence. Manifestations of repetition Repetition manifests itself in time and space. On the visual level it primarily occurs in space, whereas on the auditory level it primarily occurs in time. On the conceptual level, it seems that the manifestation of repetition takes a rather complex form as it involves at least two dimensions – mental time and space. It can also cross time or culture as historical quotation or cultural allusion on the conceptual level. As shown in Figure 2.5, repetition appears in the text in several prototypical types. A certain item may be repeated in abundance. Repetition may manifest itself in certain locations such as of regular intervals,

48

Metaphor and Iconicity

A. Repetition in quantity

B. Repetition in similar intervals

-R----------R-----------R-----------R-----------R---------------R-----------R-----------R-----------R-----------R-------R-----------R-----------R-----------R----------------R--------R-----------R-----------R----------R -----------R-----------R------------R----------R------------R---------R-----------R----------------------R------R---------------R-----------R----------R------------R------------------R-----R-----------R-------------R-------R----------R -----------R-----------R-------------------R---------------------------R-----------R---------------R--------------R---

--------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R------------------R--------------R--------------R-----------

C. Repetition at the beginning and the end

D. Repetition in sequence

R---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------R

------------------------------------RSTUVWXWZ RSTUVWXWZ -------------------------------------

E. Repetition in pairs

F. Repetition by symmetry or other orders

--------------------------R --------------------------R --------------------------S --------------------------S

--------------------------R------------------------R-------------------------------------------------R --------------------------R------------------------R-------------------------------------------------R

G. Change in repetitive units and/or in repetition patterns

H. Change to non-repetition

R-----------------R----------------R---------------R R-----------------R----------------R---------------R S--------S--------S-------S-------S-------S------S S--------S--------S-------S-------S-------S------S

R---------------R--------------R--------------R R---------------R--------------R--------------R ---------------------------------------------------R---------------R--------------R--------------R R---------------R--------------R--------------R

Figure 2.5 Manifestations of repetition types

in symmetry, in pairs, or at the beginning and the end of text. Repetitious patterns may change, from one to another, or from one to none. Even from the non-exhaustive list of abstract repetition types in Figure 2.5, it is clear that when a certain item is repeated in a certain pattern, it is likely to be perceived as ‘ﬁgure’ against ‘ground’. If the content expressed by the ﬁgured repetition conveys some meaning which ‘ﬁts’ with the ﬁgure, then the link between form and meaning is felt stronger, and becomes self-evident.

Methodological Framework 49

2.4.3 Repetition as grammatical metaphor Iconic meaning of repetition Sapir mentions that the repetition of all or part of the radical element is ‘generally employed, with self-evident symbolism, to indicate such concepts as distribution, plurality, repetition, customary activity, increase of size, added intensity, continuance’ (1921: 76). Taking up Sapir’s insight into the ‘symbolism of repetition’, Hiraga (1993) analyses Japanese and English concrete poetry to demonstrate how repetition types appear on the visual level of manifestations, and how these types of repetition convey iconic meanings respectively. It is common that a piece of poetic text manifests several repetition types at a time even when one looks at visual elements exclusively. However, the types of repetition patterns almost self-evidently show certain correspondences to the contents they express as indicated in Table 2.5. It can be assumed that this kind of taxonomy of repetition types and their corresponding iconic meanings is applicable not only at the level of visual manifestations but also at the level of auditory manifestations. Repetition navigates textual interpretation Combined with grammatical metaphors discussed in Section 2.3.3, the correspondence of repetition types and their iconic meanings gives rise to another set of grammatical metaphors, which contribute to the interplay of textual form and meaning at the level of the macro-structure of the text.

Table 2.5 Iconic meanings of repetition Repetition type

Iconic meaning

Repetition in quantity

Dominance, abundance, continuation, monotone Regularity, unity, coherence Framing, cycle Coupling, contrast Emphasis, echo Thematic development Change in content, mood, effect, etc. Foregrounding, distinctiveness

Repetition in similar intervals Repetition at the beginning and the end Repetition in pairs Repetition in sequence Repetition by symmetry or other orders Change in repetitive units and/or in repetitive patterns Change from repetition to non-repetition

50

Metaphor and Iconicity

For example, if a certain text contains a number of repetitions in pairs, they may navigate the interpretation of the text as a dominance of the concepts expressed by the paired segments in repetition, which may further signal coupling or contrast. Here, the grammatical metaphors of SIMILAR CONTENT IS SIMILAR FORM, A PAIRED CONTENT IS A PAIRED FORM and MORE CONTENT IS MORE FORM are working at two levels. At the level of interpretation of segmental elements, the similar sound sequence (FORM), for instance, may signal the similar (or equivalent) interpretation (CONTENT) as in the case of alliterations and rhymed feet. Identiﬁcation of the similarity presupposes the notion of repetition, as it is perceived that an item A is similar (or equivalent) as an item B based on the realisation that B is a repetition of A. At the level of textual interpretation, on the other hand, if a certain sound sequence is repeated many times in pairs, the abundance (MORE FORM) of such repeated pairs (PAIRED FORM) suggests a dominance (MORE CONTENT) of the paired meanings (PAIRED CONTENT). Thus, repetition can function as grammatical metaphors at the level of the macro-structure of the text, and therefore it guides or reinforces the textual interpretation as listed in Table 2.5. In terms of repetition in the text, DOMINANCE or EMPHASIS IS QUANTITY; REGULARITY, UNITY, or DEVELOPMENT IS ORDER; PAIRS MEAN COUPLING or CONTRAST; A FRAME or A CYCLE IS SIMILARITY (EQUIVALENCE) OF BEGINNING AND ENDING; and CHANGE IN CONTENT IS CHANGE OF FORM.

2.5 Summary To sum up the main argument of the methodological framework, it has been discussed that there are two types of metaphor–icon links in language: iconicity in metaphor and metaphor in icon. Iconicity in metaphor concerns the imagic and diagrammatic representation in the creation of meaning in metaphor. This is most prototypically illustrated in the notion of image-schematic structures crucial in the cognitive account for metaphor. Metaphor in icon also relates to imagic and diagrammatic aspects of the linguistic form. Conventional metaphors which conceptualise our everyday experiences and reality also conceptualise our understanding of language structure and use. These metaphors navigate the way one interprets the forms of linguistic expressions as structural or relational diagrams. The relationships of these two types of metaphor–icons link may be graphically represented by Figure 2.6. The model in Figure 2.6 is based on the Turner–Fauconnier model, but elaborated by the addition of the

Methodological Framework 51 Generic Space

Grammatical Metaphors

MEANING

FORM

Generic Space DIAGRAMMATIC MAPPING

Visual Structure Sound Structure Syntactic Structure

IMAGIC MAPPING Sounds Visual Shape

DIAGRAMMATIC MAPPING

DIAGRAMMATIC MAPPING

Visual Structure DIAGRAMMATIC Source Sound Structure MAPPING Input Syntactic Structure

Target Input

DIAGRAMMATIC MAPPING Blend

Image-Schema DIAGRAMMATIC MAPPING Image Content IMAGIC MAPPING Sensory Perception Figure 2.6 Metaphor–icon links in language12

52

Metaphor and Iconicity

metaphor–icon interplay indicated by arrows. In the box of ‘meaning’, the metaphorical process is illustrated according to the model of blending. There are four mental spaces (represented by four circles): two input spaces, a generic space and a blend. Imagic and diagrammatic iconicity relate to: (1) the inside structure of input spaces (as shown in the balloon in Figure 2.6) in which sensory perception is projected onto the image content by imagic mapping, and then onto the image-schema by diagrammatic mapping; and (2) the relationship of inputs to the generic space (inside the meaning box) in which diagrammatic mapping projects the image-schema in one space onto the image-schema in another. The box of ‘form’ represents linguistic and other semiotic resources. When the form (sounds and/or visual shape) directly mirrors the meaning, as in the case of onomatopoeia and visible language, the imagic mapping occurs in the direction from form to meaning, as illustrated by an arrow. On the other hand, when grammatical metaphors give meaning to form, there occurs a diagrammatic mapping in the direction from meaning to form (visual, sound and syntactic structures), also illustrated by an arrow. There are three types of occurrence of diagrammatic mapping: from the input space, from the generic space and from the blended space onto the form. The general structure included by these diagrammatic mappings of form and meaning constitutes the upper generic space in which the grammatical metaphors reside. Grammatical metaphors are conventional conceptual metaphors which give iconic meanings to form, and hence contribute to the interplay of form and meaning at the level of both text segments and text macrostructure. The effectiveness of this model is in the ability to specify which part or parts of the metaphorical process – whether the input, generic or blended spaces – relate(s) to the diagrammatic mapping of form and meaning. In theory, there are four mapping possibilities: from generic space onto form; from input source space to form; from input target space to form; and from blended space to form. However, in practice, the third case is unlikely because the target space in metaphor, by its own nature, is where the mapping ﬁnalises rather than initiates. The model of blending also accounts for the dynamic and creative interplay between metaphor and iconicity by way of ‘an emergent structure’ as represented by the square inside the blended space in Figure 2.6. Creation and interpretation of metaphor and icon are not a ﬁxed process, which is pre-set prior to the text. Instead, meaning of metaphor and icon is emergent at the time of the blend in which all the mental spaces are integrated, completed and elaborated into the larger system

Methodological Framework 53

by inferences, emotions, background knowledge, cultural and cognitive models. The following chapter will show that the interplay of metaphor and iconicity is manifested at the visual and the auditory levels, in various combinations. Through detailed analyses of a few prototypical examples, it will be demonstrated that the poetic texts chosen indeed foreground metaphor–icon links and thus serve as optimal data to support the theoretical claims made in this chapter.

This page intentionally left blank

Part II Analysis

This page intentionally left blank

3 Manifestation of Metaphor–Icon Links: Prototypical Examples

. . . there is clearly an iconic element in metaphor. Paul Henle (1958) The purpose of this chapter is twofold: to develop the idea that poetic texts foreground the interplay between metaphor and iconicity; and to demonstrate, with prototypical examples, the theoretical claims and methodological procedures discussed in the previous chapters. As a prototypical illustration of the complexity of metaphor–icon links, I would like to present an analysis of two short English poems and two haiku texts. These poems display both local and global metaphors. Local metaphors are placed in various parts of the poem. They are based on either conceptual mappings (in which structured image-schematic concepts are mapped from one domain onto another [involving diagrammatic iconicity]), image mappings (in which locally evoked images are mapped from one onto another [involving imagic iconicity]), or a combination of both. The local metaphors may or may not contribute to the formation of a global conceptual metaphor, which is a metaphorical reading of the whole text, i.e. the poem as metaphor (see Section 2.1.1 for details). The poems analysed below differ in the degree to which metaphor and iconicity are manifested in combination. The ﬁrst one is an example of English pattern poetry, which displays an overt imagic iconicity. By contrast, the other three poems analysed in this chapter covertly contain diagrammatic iconicity. The second example is a manifestation of diagrammatic patterns in a single poetic text where diagrammatic iconicity is achieved through the choice of lexical items and grammatical parallelism. The two haiku are chosen to show that a larger text such as a literary work containing a series of poems can also display iconic 57

58

Metaphor and Iconicity

structure as a whole. The analysis will demonstrate that in all examples, metaphor–icon links are at work strongly in the creation and interpretation of meaning. In addition, the model of blending is adduced as a means of clarifying the dynamic mechanism of metaphor–icon links, because it allows us to see which part of the metaphorical process relates to which type of iconic mapping in the poetic text.

3.1 Shape and sense in Herbert’s ‘Easter Wings’ 3.1.1 Text as shape ‘Easter Wings’ by George Herbert (1880 [1633]: 34–5) is one of the bestknown examples of pattern poetry in the English language.1 The poem is presented as a silhouette of wings. The visual image mirrors the theme of the poem, which is concisely stated in its title, ‘Easter Wings’. This is a typical example of imagic iconicity, as the relationship of form and meaning is immediate and mimetic. The very shape of the poem – the rising wings of two birds – exhibits the content of the poem in which a human being relates to the actions of a bird (speciﬁcally as a lark). Example 3.1

Manifestation of Metaphor–Icon Links 59

N.B. The following ‘prosaic’ presentation of the poem is provided only for the clariﬁcation of the content of this graphic representation: Lord, who createdst man in wealth and store, Though foolishly he lost the same, Decaying more and more, Till he became Most poore: With thee O let me rise As larks, harmoniously, And sing this day thy victories: Then shall the fall further the ﬂight in me. My tender age in sorrow did beginne: And still with sicknesses and shame Thou didst so punish sinne, That I became Most thinne. With thee Let me combine And feel this day thy victorie: For, if I imp my wing on thine, Afﬂiction shall advance the ﬂight in me. 3.1.2 Text as icon Overall metaphor–icon links There are at least two major input spaces created by the reading of this poem as a global metaphor: a bird and a human being. This is done not only in linguistic expressions such as ‘let me rise as larks’, ‘the ﬂight in me’, and ‘imp my wing on thine’, but also in graphic expressions such as the shape of the two sets of wings, and their layout on two facing pages. In the ﬁrst stanza, a human being (both ‘Adam’ and all humanity) is (re)presented as a bird which is about to rise on Easter Day. God created Adam (‘man’) giving him wealth, but he fell and forfeited his paradise. Man’s abundance is steadily ‘decaying’ into poverty. God’s only son, Christ (‘thee’), enables man to ‘rise’ again spiritually just as Christ rose from the dead on Easter Day. In the second stanza, with the inheritance of original sin, humanity begins in ‘sorrow’, and also decays into sickness. As in the ﬁrst stanza, it is ‘with thee’ that provides an

60

Metaphor and Iconicity

Iconic Mapping

Metaphorical Mapping

FORM

MEANING Generic Space

BIRD WINGS

IMAGIC MAPPING

HUMAN

BIRD WINGS

DIAGRAMMATIC MAPPING

Blend

Grammatical Metaphors Generic Space Figure 3.1 Metaphor–icon links in ‘Easter Wings’

uplift for humanity and for the individual. The narrator sees himself as a bird which will ‘imp’ his wing with feathers from Christ’s wings so as to make the ‘ﬂight’ to heaven. In effect, both the physical ﬂight (or strength to go on with life) and the spiritual uplift are implicated in the poem. As illustrated in Figure 3.1, a bird is both the input source space for a global metaphorical mapping (onto a human being) and the target space for an imagic iconic mapping (from the visual source space of the wings of a bird). The metaphorical mapping of a bird onto a human being is represented by the four mental spaces on the right-hand side of Figure 3.1. The source input space has an image-schema of a bird – a lark with

Manifestation of Metaphor–Icon Links 61

wings, singing, ﬂying, and the movement (expansion and contraction) of the wings. The target input space, on the other hand, has an imageschema of a human being – Adam (‘man’), a narrator (human being in general), sin and retribution. In the blended space, a fused imageschema of a bird and a human being receives a dynamic structure from the cultural background of Christianity. Adam’s original sin is related to the punishment, mortality and death of human beings. Man’s wealth is shrinking and decaying into poverty. The soul is also in decline. Then, Christ (as represented by ‘thee’) enables man to rise again spiritually. The victories of Christ on Easter Day are celebrated by the song of a lark. The movement of wings, i.e. expansion and contraction, is mapped onto the processes of salvation and decay, respectively, as they relate to human beings. This metaphorical reading leads on to suggest a diagrammatic iconic mapping between the form (the space on the left-hand side) and the blended space, as indicated by the arrow in Figure 3.1. It should be noted that the wings, an icon for the bird, had to be put vertically on the pages in order to display the fact that they are wings and the bird is rising. The wing shape is an imagic icon because it is a direct mimicry. The rising shape, on the other hand, is a result of the reading of the poem’s content. The wings on the second page look smaller because the ﬁrst and the last lines are shorter than those on the previous page. They give the impression that the bird on the second page has taken off or is ﬂying further. Seen as a global metaphor, the vertical form of the two wings is charged by layers of religious readings provided by Christian tradition and culture. The form embodies the metaphorical meaning of Easter as celebrating Christ’s being born again, or rising from death. The form also awakens associations with splendidly winged angels, and image patterns such as fall and rise, decay and advance, loss and victory, to list but a few. Diagrammatic iconicity by grammatical metaphors The poem, moreover, exhibits structural correspondences between form and meaning navigated by conventional metaphors for linguistic forms as discussed in the previous chapter (see Sections 2.3.3 and 2.4.3 in particular). The narrowest lines are given emphasis both by location, i.e. at the centre of each stanza, and length, i.e. their brevity. The central position of the wings displays a phrase repeated in both verses of the poem, ‘With thee’. If one interprets this poem as a religious declaration against a Christian cultural background, ‘with thee’ is central to the poem’s

62

Metaphor and Iconicity

declaration of belief. The focal signiﬁcance of ‘with thee’ is centralised visually or physically in the core of the poem. In addition, ‘with thee’ at the joint of the wings suggests a relationship – as if that most poor creature at his most diminished is saved by a hand reached out. The other short lines, one in each stanza, ‘most poor’ and ‘most thin’, form a pair. There is an analogical link between the shortness of form and the smallness of the objects (‘man’ in the ﬁrst stanza, and ‘I’ in the second stanza) described.2 The third line in the ﬁrst stanza, ‘Decaying more and more’, also corresponds diagrammatically to the decrease in the length of the lines of the poem. These correspondences are more subtle and less noticeable than the overall silhouette of the poem; nonetheless, they are as important. These are cases of diagrammatic iconicity. Imagic iconicity by visual layout It is clear that the imagic iconicity dominates in this text because, if we put the words of the poem in an ordinary prosaic format or even in a horizontal layout as in Example 3.2, the intended iconic effect of the poem is easily lost (Harris 1995: 61).3 Example 3.2 Lord, who createdst man in wealth and store, Though foolishly he lost the same, Decaying more and more, Till he became Most poore: With thee O let me rise As larks, harmoniously, And sing this day thy victories: Then shall the fall further the ﬂight in me. My tender age in sorrow did beginne: And still with sicknesses and shame Thou didst so punish sinne, That I became Most thinne. With thee Let me combine And feel this day thy victorie: For, if I imp my wing on thine, Afﬂiction shall advance the ﬂight in me.

Manifestation of Metaphor–Icon Links 63

In short, the meaning expressed by the language and the visual form of this poem constitute simultaneously an image, a diagram and a metaphor of ‘Easter Wings’. The image is the dominant iconic manifestation in this particular example; but the creativity of this image is enhanced by the diagrammatic and metaphorical manifestations as clariﬁed by the analysis.

3.2 Grammatical parallelism in Shelley’s ‘Love’s Philosophy’ 3.2.1 Text as metaphor Example 3.3 Love’s Philosophy4 The fountains mingle with the river, And the rivers with the ocean, The winds of heaven mix for ever With a sweet emotion; Nothing in the world is single, All things by a law divine In one another’s being mingle – Why not I with thine? See the mountains kiss high heaven, And the waves clasp one another; No sister ﬂower would be forgiven If it disdained its brother: And the sunlight clasps the earth, And the moonbeams kiss the sea – What are all these kissings worth, If thou kiss not me?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Local metaphors P. B. Shelley’s ‘Love’s Philosophy’ (1904 [1820]: 216) abounds in the use of local metaphors.5 The following analysis will clarify what conventional metaphors are employed and how they are sculpted into the composite metaphorical conception for the given target space, which is hinted at by the title of the poem, ‘Love’s Philosophy’. The abundant use of metaphorical expressions reﬂects conventional cognitive metaphors as follows in Example 3.4.

64

Metaphor and Iconicity

Example 3.4 (a) A NATURAL PHENOMENON IS AN ENTITY (b) A NATURAL OBJECT IS AN ENTITY (c) FEELING IS AN ENTITY (d) NATURE IS HUMAN (e) OPPOSITES FORM PAIRS (f) FUSION IS MIXING (g) UNDERSTANDING IS SEEING (h) CONTACT IS TOUCHING (i) REQUESTING IS QUESTIONING (j) LOVE IS A FUSION OF OPPOSITES (k) LOVE IS CONTACT (l) LOVE IS PAIRING The ﬁrst cognitive metaphors, A NATURAL PHENOMENON IS AN ENTITY, A and FEELING IS AN ENTITY, are what Lakoff and Johnson (1980: 25–32) call ‘ontological metaphors’ which enable us to conceive continuous events, ongoing realities, and unstable or even moving states of things as if they were quantiﬁable objects, entities or units with which we can refer to, quantify, identify and so on. In this poem, NATURAL PHENOMENA and NATURAL OBJECTS such as ‘fountains’, ‘rivers’, ‘winds’, ‘mountains’, ‘heaven’, ‘waves’, ‘sunlight’, ‘earth’, ‘moonbeams’ and ‘sea’ are treated as discrete ENTITIES which have boundaries and shapes. ‘Emotion’ is seen as if it were an ENTITY, which is countable and even modiﬁable by an adjective of taste, ‘sweet’. Furthermore, some of these entities are personiﬁed by the NATURE IS HUMAN metaphor: e.g. the ‘mountains’ that ‘kiss’ high ‘heaven’; the ‘waves’ that ‘clasp’ one another; the ‘sister ﬂower’ that ‘disdained’ its ‘brother ﬂower’; the ‘sunlight’ that clasps the ‘earth’; and the ‘moonbeams’ that ‘kiss’ the ‘sea’. It might be worth pointing out in this connection that some of the metaphors are what Lakoff and Turner (1989) classiﬁed as a non-conventional expansion of conventional metaphors. For example, the metaphor of a ‘sister-ﬂower’ disdaining its ‘brotherﬂower’ in lines 11 and 12 extends the NATURE IS HUMAN metaphor to include not only the male/female distinction but also the particular kinship of brother and sister. The ENTITIES are put into semantic oppositions such as: LARGE vs. SMALL (‘river’ vs. ‘fountain’; ‘ocean’ vs. ‘river’); HUMAN vs. DIVINE (‘emotion’ vs. ‘heaven’); HUMAN vs. NATURE (‘emotion’, ‘I’, ‘thou’, ‘me’, ‘thine’, ‘brother’, ‘sister’ vs. ‘fountains’, ‘river’, ‘ocean’, ‘wind’, ‘mountains’, ‘waves’, ‘ﬂower’, ‘sunlight’, ‘earth’, ‘moonbeams’, ‘sea’); EARTH vs. NATURAL OBJECT IS AN ENTITY

Manifestation of Metaphor–Icon Links 65 HEAVEN (‘fountains’, ‘river’, ‘ocean’, ‘mountains’, ‘waves’, ‘ﬂower’, ‘earth’, ‘sea’ vs. ‘heaven’, ‘law’, ‘divine’, ‘sunlight’, ‘moonbeams’); MAN vs. WOMAN (‘brother’ vs. ‘sister’; ‘I’ vs. ‘thou’; ‘me’ vs. ‘thine’); WATER vs. SOIL (‘fountains’, ‘river’, ‘ocean’, ‘waves’, ‘sea’ vs. ‘earth’, ‘mountains’); and DAY vs. NIGHT (‘sunlight’ vs. ‘moonbeams’). FUSION is expressed by the verbs of MIXING such as ‘mingle’ and ‘mix’. What is FUSED has the same quality: SHAPELESSNESS. The image of FUSION is thus introduced by the MIXTURE OF THE FLOW OF LIQUID, such as the ‘fountains’ mixed with the ‘rivers’, and the ‘rivers’ with the ‘ocean’, and by the MIXTURE OF THE FLOW OF THE INVISIBLE, such as the ‘winds’ mixed with each other with a ‘sweet emotion’. As the poem progresses, the image of FUSION turns into the image of CONTACT, particularly PHYSICAL CONTACT. It seems to indicate the progression of the lover’s longing from an abstract fantasy to a concrete vision. The opening line of the second stanza starts with the verb ‘see’, in the imperative mood. This is a focal point. The act of seeing merges with its double meaning via the metaphor UNDERSTANDING IS SEEING, and the seeing by the physical eye and the mind’s eye fuse. The cognitive metaphor UNDERSTANDING IS SEEING is pervasive in ordinary language. For example, ‘I see what you mean’, ‘It looks different from my point of view’, ‘I was totally in the dark during the presentation she gave’, ‘I’ve got the whole picture’, and ‘He gave an illuminating paper’. CONTACT is expressed by the verbs of PHYSICAL BINDING and TOUCHING such as ‘clasp’ and ‘kiss’. Those being united by PHYSICAL CONTACT are PERSONS, i.e. PERSONIFIED NATURAL OBJECTS. Obviously the poem uses nonconventional elaboration of image-schemas such as ‘the mountains kiss the heaven’, ‘the sunlight clasps the earth’, ‘the moonbeam kisses the sea’. In the conventional metaphors, PERSONIFIED NATURE may ‘touch’ and ‘manipulate’ objects such as ‘the mountain ridge touched the sky’, or ‘wind carries seeds’; but it does not interact with them to the extent that this poem describes. PAIRING is implied by the semantic oppositions as pointed out above; but it is expressed in more speciﬁc terms by a paired set of antonyms such as ‘I’ and ‘thou’, ‘brother’ and ‘sister’, ‘sun’ and ‘moon’. The image of PAIRING is also supported by the abundance of MIXING and CONTACT metaphors. Each stanza ends with a question, which also implies a double meaning: an inquiry and a request. While the questions at the end of the poem are direct questions, they also function as a request, longing and/or wooing (indirect speech acts). Thus, each stanza is a REQUEST (LONGING, WOOING) in the form of a QUESTION.

66

Metaphor and Iconicity

Global metaphor Now that it is clear that many of the metaphorical expressions in this poem are based on the conventional cognitive metaphors that we use in everyday language, then the question is how Shelley exploited conventional metaphors to create his unique poetic text. As has been pointed out, he used a non-conventional extension and elaboration of conventional metaphors to some extent. Nevertheless, the most prominent technique is the formation of composite metaphors by combining multiple conventional metaphors together to produce a richer and more complex set of metaphorical networks, which gives inferences beyond those that follow from each of the metaphors alone. The overall global mapping present in the poem is a mapping of the NATURE domain onto the HUMAN domain. All the MIXING and FUSION and CONTACT in NATURE are mapped onto the PAIRING of two lovers. The network of composite metaphors discussed above constitutes a series of local blends, which are recruited at the time of the global blend of NATURE and HUMANS. The input source space has a rich image-schematic structure provided by various pairing acts of natural phenomena and objects. The input target space recruits an image-schematic structure from our common-sense knowledge and background information about the lovers. The image-schematic structure of each input space is preserved in the abstract structure in the generic space, a frame of PAIRING, in this example. The frame of PAIRING is schematised in two ways: a coupling of the two similar things, on the one hand (e.g. the water and the water, the waves and the waves, the ﬂower and the ﬂower, the human and the human), and of two different things, on the other (e.g. what is solid and what is air, brother and sister, man and woman). In the blended space, all these structures from the generic space and the input spaces are preserved, and in addition, some emergent structures such as the separation of the two lovers, arise. In spite of the fact that the metaphors in the poem concern the mapping of mixing, fusion and contrast found in nature, onto the pairing of two lovers, the lovers themselves are not united. Shelley’s poem is a blend which recruits inferences not only from linguistic conventions instantiated in a composite metaphorical network but also from common-sense knowledge, culture, tradition, history and religion. The poem can be a general statement about love’s philosophy. It can also address the potentiality of what may happen. Or it can also be Shelley’s own message to his lover. The poetic imagination extends the conventional metaphors of love to

Manifestation of Metaphor–Icon Links 67

the realm both of the poet’s particular experience and of the potential experience of all lovers. All reside in the blend. 3.2.2 Iconic interpretation by metaphor Figure 3.2 illustrates the metaphorical blend discussed above and the diagrammatic iconic mappings to be explained in detail below.

Iconic Mapping

Generic Space 2

Metaphorical Mapping

Grammatical Metaphors PAIR

FORM

MEANING

Generic Space 1 PAIR

DIAGRAMMATIC MAPPING

PAIR

OPPOSITION DIAGRAMMATIC MAPPING

NATURE

HUMAN

Blend

OPPOSITION Grammatical Generic Space 3 Metaphors Figure 3.2 Metaphor–icon links in ‘Love’s Philosophy’

68

Metaphor and Iconicity

The point in this particular poetic text is that the poem’s form can acquire an iconic meaning by the grammatical metaphors which actually constitute the content of the two generic spaces (Generic Spaces 2 and 3) for the two iconic mappings. Firstly, there is a diagrammatic mapping between the schematic structural form of the poem and the image-schema of Generic Space 1, which is a frame pairing the similar and the different. This mapping is supported by the grammatical metaphor in the Generic Space 2, A PAIRED CONTENT IS A PAIRED FORM. Secondly, there is another diagrammatic mapping between the schematic structural form of the poem and the image-schema of the blended space. This concerns an iconic mapping of the image-schema of separation projected through Generic Space 3, AN OPPOSING CONTENT IS AN OPPOSING FORM. A PAIRED CONTENT IS A PAIRED FORM

In terms of the lexical architecture of the poem, the frequency of the use of certain words in this poem mirrors PAIRING, which is a shared image-schematic frame of the global metaphor. Even a hasty reader would notice that several lexical items are used repeatedly. The most prominent characteristic of these repetitions is that the same (or similar) items are repeated only twice as if they are PAIRING each other. There are ten identical pairs and ﬁve semi-identical pairs, which are either derivationally or inﬂectionally similar. Table 3.1 shows these pairs and their position in the text. On the stanzaic level, too, the ﬁrst and the second stanza form a PAIR in such a way that each stanza has the same number of elements lexically, syntactically and prosodically, as summarised in Table 3.2. Lexically, the ﬁrst and the second stanza have the following: six the’s, two be’s (including ‘being’), one lexical item repeated four times (i.e. four with’s in the ﬁrst stanza and four kiss’s in the second), four adjectives (i.e. adjectives – sweet, single, all and divine in the ﬁrst stanza and high, sister, all and worth in the second), four plural nouns (i.e. fountains, rivers, winds and things in the ﬁrst stanza and mountains, waves, moonbeams and kissings in the second). Syntactically, there are ﬁve afﬁrmative sentences and one interrogative sentence in each stanza. And prosodically, both stanzas have the same end rhyme scheme, i.e. ababcdcd. AN OPPOSING CONTENT IS AN OPPOSING FORM

The fact that the poem consists of two stanzas indicates that the stanzaic form of the poem itself is a means of the OPPOSING PAIRS making

Manifestation of Metaphor–Icon Links 69 Table 3.1 Lexical pairs in ‘Love’s Philosophy’ Parts of speech

Lexical item

Line number

Verb

mingle clasp river heaven one another I – me thine – thou it – its all a no – nothing why – what not in if

1 and 7 10 and 13 1 and 2 3 and 9 7 and 10 8 and 16 8 and 16 12 and 12 6 and 15 4 and 6 5 and 11 8 and 15 8 and 16 5 and 7 12 and 16

Noun Pronoun

Determiner

Adverb Preposition Conjunction

Table 3.2 Stanzaic pairs in ‘Love’s Philosophy’

Lexical items repeatedly used Adjectives Plural nouns

Introductory sentence Concluding sentence Rhyme scheme

First stanza

Second stanza

six the’s two be’s four with’s four adjectives – sweet, single, all and divine four nouns – fountains, rivers, winds and things afﬁrmative interrogative ababcdcd

six the’s two be’s four kiss’s four adjectives – high, sister,6 all and worth four nouns – mountains, waves, moonbeams and kissings afﬁrmative interrogative ababcdcd

CONTACT. This also holds true in the alternation of indentation of the even lines, with further alternation of paired line initials (And–With– All–Why in the ﬁrst stanza, and And–If–And–If in the second stanza). The syntactic features of some lexical elements and the syntactic patterns in the text also reﬂect the image-schemas of OPPOSITES. Firstly, the OPPOSITION of ‘I’ and ‘thou’ seems to be supported by the presupposed SEPARATION between ‘I’ and ‘thou’, expressed by the ﬁnal questions of

70

Metaphor and Iconicity

each stanza, which contain a negation: ‘why not I with thine?’ and ‘what are all these kissings worth, if thou kiss not me?’ (My italics.) Secondly, the OPPOSITIONS are also embodied in the grammatical differences between the ﬁrst and the second stanza. In the ﬁrst stanza, the verbs are all intransitive, and each line has one preposition, whereas in the second stanza, the verbs are all transitive, and there is no preposition. Hence, the ﬁrst stanza and the second stanza are grammatically contrastive. In addition, the afﬁrmative introductory sentences and the interrogative concluding sentence form an OPPOSITION within each stanza, though they form a PAIR between the ﬁrst and the second stanza. Multi-layered metaphor–icon links It has been seen so far how this poem manifests multi-layered metaphor–icon links. There are several conventional conceptual metaphors sculpted into the composite metaphorical conception of LOVE. There are images in local metaphors, too, which enrich the feel of the poem. A mixing act of waters and of invisible air is mapped onto a mingling act of lovers. The contact of visible natural objects and of fragrant ﬂowers is mapped onto the kissing of lovers. The lexical, syntactic and stanzaic form of the poem displays a diagrammatic relationship to the content of the text navigated by the global metaphorical blend as well as conventional grammatical metaphors. To summarise, ‘Love’s Philosophy’, describing the longing of a narrator to be one with his separated (or unreachable) lover, displays the theme iconically and metaphorically in its cognitive content, linguistic form and image, particularly achieved by a creative choice of vocabulary and grammatical patterns. Indeed, the poem is a metaphorical icon of ‘love’s philosophy’. The analysis by the model of blending has demonstrated how and where the organic network of the poem is buttressed by diagrammatic and metaphorical coherence between the formal dimension and the dimension of concept and image in the poetic text.

3.3 Isomorphism of text structure in Basho’s Oku no Hosomichi 3.3.1 Haiku and cognitive metaphor Data selection The interplay of metaphor and iconicity is not only manifested at the level of a single poem but also at the level of a whole literary text. In

Manifestation of Metaphor–Icon Links 71

this section, I will look at two haiku texts taken from Oku no Hosomichi (Matsuo 1996[1694]). Oku no Hosomichi was written as a travel sketch, which consisted of a main narrative body, 50 haiku poems by Basho, and a few other poems by other authors. The 50 haiku poems are considered as an integrated text in its own right, conforming to the general principle of composition and structural congruence. The two haiku texts were chosen because structurally they form a thematic pair that frames the larger text of the travel sketch, the ﬁrst poem in the sketch’s opening phase, and the second at its conclusion. In my analysis, I hope to demonstrate that metaphor–icon links manifest themselves not only at the level of respective haiku texts but also at the level of macro-structure of the entire literary work of the travel sketch. In other words, it is an attempt to substantiate the claim for the effectiveness of the cognitive approach made by literary critics such as Donald Freeman (1993: 1): ‘cognitive metaphor provides accounts of language patterns that are isomorphic with larger imaginative literary structures, as well as particular interpretations that are more explicit and falsiﬁable than existing interpretations founded upon the language of literary works’. According to the model of blending, cognitive metaphors, both local and global, will be analysed in conjunction with the iconicity they yield. Blending and haiku text The blending model appears to be particularly useful in analysing haiku texts for the following three reasons: 1. It stresses the importance of ‘the emergent structure’ of the blended space, and therefore provides an effective tool for understanding the creativity of literary metaphors (not only of haiku but also of any poetic text); 2. Blending better explains than the past literary studies (cf. Kobayashi 1966, Toyama 1970) the rhetorical effects of loose grammatical conﬁgurations in the haiku texts such as the juxtaposition of phrases by kireji (‘cutting letters’ or rhetorical punctuation) and kake-kotoba (‘hanging words’ or puns); 3. The blending approach, which does not specify unidirectional mapping between input spaces, seems to provide a better means of understanding the frequent occurrence of allegory7 and personiﬁcation in haiku than the standard two-domain model of cognitive metaphor (cf. Lakoff and Johnson 1980, Lakoff 1987, 1993, Lakoff and Turner 1989).

72

Metaphor and Iconicity

Cultural cognitive model It is also pointed out that understanding haiku texts, which are extremely short in form and rich in traditional implications, requires a deep understanding of traditional Japanese culture, which shapes the cultural cognitive model. A non-exhaustive list of the features of such an understanding includes: the actual context such as time, place, customs, life and so on, which contextualise the poem in general terms; folk models, which originate from myth, folk beliefs and concepts of the time; conventional metaphors (cf. Lakoff and Johnson 1980), which have been adopted into a given speech community over time, and which a poet can exploit in non-conventional ways; and the iconicity of kanji, Chinese logographs, which link form and meaning, particularly with regard to their etymological derivation, and thereby serve as a cognitive medium for haiku texts. As stated before, Japanese has three different writing systems: kanji (to write words of Chinese origin), and the two types of Japanese syllabary, hiragana (to write words of Japanese origin, most function words and conjugations) and katakana (to write words of foreign origin other than Chinese). The complexity of the Japanese writing systems gives a unique dimension to a visual and conceptual representation of haiku. The following analysis will look at how such knowledge is activated and recruited when reading the poems. 3.3.2 Texts Before looking at the macro-structure of the entire text, it is necessary to explain the two haiku poems, which play an important role in structuring the travel sketch. As mentioned earlier, these poems were placed at the beginning phase and in the conclusion of the travel sketch. Example 3.5 (a) Yuku haru ya going spring:

tori naki uo no birds cry ﬁsh of

me wa namida eyes TOP tears8

‘Departing spring: birds cry and, in the eyes of ﬁsh, tears’ (Matsuo 1996 [1694]: 43).9 (b) Hamaguri no a clam GEN/SUB

futami ni wakare lid. ﬂesh into separate Futami (place name) to separate

yuku aki zo going autumn INT

Manifestation of Metaphor–Icon Links 73

‘A clam separates lid from ﬂesh as autumn departs’ (Matsuo 1996 [1694]: 133). These two haiku texts both describe a moment of sadness caused by departure, one at the start of a trip and the other at the end. Both poems further depict sadness in images of separation and transience, which are embodied in metaphors of natural species and of the passing seasons. Speciﬁcally, crying birds and ﬁsh suggest tears of sadness; the clam-shell separated from its ﬂesh stands for the pain of leave-taking. The passing of spring and autumn are tropes that symbolise the transience of nature, which in turn causes pangs of sorrow. ‘Spring’ and ‘autumn’ are natural events that here represent human action, namely that of a traveller who is ‘departing’. Birds, ﬁsh and a clam are also personiﬁed in the text in that birds and ﬁsh cry with tears, while the clam-shell and the clam-ﬂesh separate. Three prominent cognitive metaphors are illustrated.10 They are TIME IS A TRAVELLER (based on EVENTS ARE ACTIONS); NATURE IS HUMAN; and its inverse version, PEOPLE ARE NATURE (ANIMALS, BIRDS, FISH, INSECTS, PLANTS, etc.). From the intra-textual context, most readers of haiku would realise that the poems do not just describe natural events, but also reﬂect the feelings of the poet and his friends as represented by birds, ﬁsh and the clam. In examining these two poems, two aspects are considered: how each haiku text manifests local and global metaphors, which are interpreted with common prior knowledge and the principle of iconicity; and how metaphor–icon links in each haiku constitute another metaphor–icon link in the larger structure of an entire literary text of the travel sketch.

3.3.3 The spring poem Overview of the spring poem Preceding the spring poem, Basho wrote: . . . My close friends, who had been gathered since the previous evening, sent me off in a boat. When we climbed out of the boat at a place called Senju, I was depressed by the thought of the three thousand li that lay ahead and shed tears at a parting in this illusory world (italics in the original). Yuku haru ya going spring:

tori naki uo no birds cry ﬁsh GEN

me wa namida eyes TOP tears

‘Departing spring: birds cry and, in the eyes of ﬁsh, tears’ (Matsuo 1996 [1694]: 43).

74

Metaphor and Iconicity

The poem, at ﬁrst reading, describes natural events: that the spring season is passing, birds are crying, and the ﬁsh have watery eyes. These natural events can then be personiﬁed, based on the metaphor NATURE IS HUMAN. Depending on the reader’s interpretation of which natural event is projected onto which human action, the poem can be read in several ways. For example, ‘passing spring’ can be interpreted as literal or metaphorical (representing, say, Basho’s departure), and the crying birds and ﬁsh as either Basho himself or his friends. At a more sophisticated or abstract level, the reader might also attribute to this poem a global metaphorical structure, in that it captures the oneness of human beings and nature, or the undifferentiated nature of human feelings and the feelings evoked by natural events, objects and creatures. This level of allegorical interpretation is based on the metaphor PEOPLE ARE NATURE, and is very much an embodiment of Basho’s poetics.11 Such non-differentiation can probably be better captured by understanding the poem as a global metaphor derived from a global blend, as illustrated in Figure 3.3. Generic Space sadness

NATURE IS HUMAN Personification HUMAN in general human beings traveller

weeping fish crying birds passing spring spring

Particular PEOPLE Basho friends

NATURE

PEOPLE ARE NATURE Allegory

Blend unity of feelings Figure 3.3 Blend of the spring poem

Manifestation of Metaphor–Icon Links 75

There are several input spaces – the passing spring season, crying birds and weeping ﬁsh as the NATURE inputs, and the human beings in general, the traveller, Basho and his friends as the HUMAN inputs. The two arrows indicate personiﬁcation and allegorical interpretation. On the one hand, the natural season and the natural creatures are represented metaphorically as a traveller and human beings respectively. On the other hand, these personiﬁed natural events and creatures are interpreted allegorically as Basho and his friends. The blend inherits a partial structure from all of the inputs; but it also has an emergent structure of its own – the revelation of the unity of feelings in nature and humans. There is also an abstract generic space, which reﬂects event frames for sadness common to both input spaces – agent, action, mental state of being sad, and so on. The global blend of the text as a whole takes, from the target input space, the feeling of sadness and regret, and the sentiment of farewell at the departure of the trip, and, from the input source space, audible wailing and visible tears of creatures.12 Each of the input spaces further constitutes a blended space of its own. Each input space is a local blend of ‘passing spring’, ‘birds crying’ and ‘ﬁsh weeping’. Inferences derived from these local blends, in addition to common prior knowledge, contribute to the understanding of this poem. Local metaphors in the spring poem The ﬁrst line of the ﬁrst poem, yuku haru (‘passing spring’), is based on the conventional metaphor, TIME (SEASON) MOVES. There are many similar examples in idiomatic Japanese: kisetsu ga meguru (‘seasons turn around’), kisetsu ga kawaru (‘seasons change’), haru ga chikazuku (‘spring is approaching’), fuyu ga kita (‘winter has come’), natsu ga sugiru (‘summer passes’), and so on. By using the verb yuku (‘to go’), the poem suggests that we are living in the present, and can feel or even see moments of spring passing us by, going away from us. In addition to this TIME (SEASON) MOVES metaphor, TIME and SEASON are also often conceptualised metaphorically in Japanese as A PERSON, especially A TRAVELLER. When people say, haru no otozure o matsu (‘to wait for the visit of spring’), aki no ashioto ga kikoeru (‘to hear the footsteps of autumn’), kibishii fuyu ga tourai-suru (‘severe winter has arrived’), for example, they are personalising the seasons. The verb yuku (‘to go’) also implies metaphorically that the motion of ‘going’ is a departure or farewell. Japan has four distinct seasons and spring is always the season of new beginnings. After the harsh winter, spring brings warmth, birth and new life. In folklore as well as today, SPRING is conceptualised as the starting

76

Metaphor and Iconicity

point of progress and prosperity, and the other seasons follow on. This is clearly reﬂected in the LIFETIME IS A YEAR metaphor (e.g. jinsei no haru ‘spring in life’, kare no jigyoo ni wa aki-kaze ga fui-te iru ‘there is autumn wind blowing against his business’). Choosing spring as the time to begin one’s long journey, therefore, accords with both its ﬁgurative and literal (real-life) implications. But why does the poet choose the end of spring? In Japan, the freshness of spring fades with the approach of a wet and humid rainy season, before the onset of summer. Hence, the last phase of spring evokes sadness or melancholy. The verb yuku (‘to go’)13 in yuku haru ya, is thus a metaphor of FAREWELL as a part of TRAVELLING. The blend of ‘the passing spring’ exploits and develops counterpart connections between TRAVELLER, the source space and SPRING, the target space. It is important to note that the poem establishes a correspondence between the departure of a TRAVELLER and the ending of SPRING, and not between the more structurally matched counterparts, of a TRAVELLER reaching a goal and the ending of SPRING. This correspondence can be guided by pragmatic knowledge recruited in the source that the poet is departing on a journey, and in the target that it took place in the last phase of spring. It is also provided more directly by common folk belief in the cyclic nature of life and time. People in Basho’s times generally believed that a death was a birth into another world; an end was a beginning of a new event. This cyclic interpretation of life and death is an underlying philosophy of Basho’s Oku no Hosomichi, which will be discussed in length below (see Section 3.3.5). Neither of the input spaces, SPRING or a TRAVELLER, necessarily implies ‘sadness’. But it is the blend of a departure fused with the ending of spring, that leads to an interpretation of the sadness of farewell as identical to the sadness of that season’s fading. Historical and contextual knowledge further suggests that the traveller was Basho, and so leads to an allegorical interpretation that Basho is represented by SPRING as A TRAVELLER. Note that if spring is a metaphor for Basho, then the source and target are related inversely. In the blend yuku haru ya, spring is at once a generic traveller and Basho in particular. A further interesting aspect of the relationship between SPRING and TRAVELLER, or more accurately a difference between the two, is that natural events like spring passing to summer cannot be stopped by human will, whereas human actions like travelling can be stopped at any time. This difference can perhaps be read as a similarity in the blend ‘the passing spring’. Since no one can stop the course of the seasons, so Basho’s trip(s) cannot be discouraged or stopped by anyone or anything.

Manifestation of Metaphor–Icon Links 77

The second line of the ﬁrst poem contains another local blend, tori naki (‘birds cry’), which can be read either be a literal expression or a metaphorical one (personiﬁcation of birds). What is interesting here and characteristic of the poem is that the possibility of multiple interpretations is reinforced by the choice of a particular Chinese logograph for naki (‘cry’), , instead of the ones more commonly used, and . Among the more common logographs, the former, , means that humans shed tears, and the latter, , means that birds, animals and insects cry aloud. The meaning of these two logographs is unequivocal as the radicals of each logograph, particularly the left-hand radical of ‘water’ for the former and the right-hand radical of ‘bird’ for the latter, contribute to specifying, rather than broadening, the meaning. By contrast, Basho’s choice of the Chinese logograph, , for this poem, suggests that he deliberately used the etymological implication and the equivocal nature of this logograph. It consists of two radicals: that on the left, , etymologically means ‘a mouth as a metonymy for voice’, while that on the right, , means ‘to wring something (usually wet)’. Thus, the logograph itself can be seen as a blend of two inputs, corresponding to the two radicals.14 This blend produces a meaning of the logograph as ‘crying in a voice produced by wringing the throat’. Furthermore, because the left-hand radical of ‘a mouth’ is ambiguous, implying both human and non-human agents, the logograph can be seen as a blend of two lexical meanings: for humans to shed tears and cry aloud in such a voice, to wail with pain; and for birds, animals and insects to cry aloud, to wail. Hence, the blend ‘birds cry’ displays a double image: birds crying aloud and humans wailing and shedding tears. Two metaphoric understandings of tori naki (‘birds cry’) are dependent on the conventional conceptual metaphors, BIRDS ARE HUMAN and its inverse version, PEOPLE ARE BIRDS. In the ﬁrst case, birds are personiﬁed in a general sense, the actions of birds are understood in terms of our own actions. In Example 3.5(a), the birds cry aloud in sorrow is an image-mapping of the crying of human beings onto the crying of the birds. Note that it is not speciﬁed whether the noun tori (‘bird(s)’) is singular or plural, nor is the type of bird speciﬁed – there is no mention of size, shape, colour or name. It is a bird or birds in a very generic sense, which in turn could be a symbol for a creature representing the sky in myth and folklore. Hence, tori naki could imply metonymically that the sky shows grief. Speciﬁcation of the bird(s) in this poem comes only from a personiﬁed action – making the noise of humans crying aloud in grief, where cry means both shedding tears and wailing. But

78

Metaphor and Iconicity

still, tori naki can be taken as describing the action of birds rather than people. Second, there is an inverse metaphorical reading, PEOPLE ARE BIRDS, which leads to an allegorical interpretation. Many proverbs express human actions as bird actions: e.g. tobu tori ato o nigosazu (‘departing birds clean up the place before’); kago no naka no tori (‘a bird in the cage’), and so on. Therefore, the personiﬁed birds in the poem, which are crying in sadness, could signify either Basho, his friends, or both. In other words, this allegorical reading comes from the historical fact which comprises the pragmatic knowledge of this poem, that the sadly crying agent could be either Basho or his friends, or could even include both of them. Thus, the equivocality of tori naki seems to be embedded iconically in the Chinese logograph of naku, , and the two-way mapping in the blend of the input spaces, HUMAN and BIRD. Figure 3.4 is a graphic representation of the interplay between metaphor and iconicity of tori naki. Firstly, the Chinese logograph, , is a blend of (‘mouth’, metonymy for voice) and (‘to wring something’, meaning of crying) represented in the box of ‘form’. Secondly, the two-way metaphorical mapping, i.e. BIRDS ARE HUMAN (personiﬁcation) and PEOPLE ARE BIRDS (allegory), in the blend of tori naki is illustrated by the two arrows in opposite directions in the box of ‘meaning’. The principle of iconicity is at work on two levels. On the level of imagic mapping, the logograph, , mirrors the blend space of crying by humans and non-humans in a speciﬁc manner. The radical, (‘mouth’), also signals towards the generic space as it is a metonymy for voice in a general and a skeletal sense. On the level of diagrammatic mapping, the equivocality of tori naki corresponds to the ambiguity of (‘crying by humans’ and ‘crying by non-humans’), navigated by the grammatical metaphor EQUIVOCALITY IN MEANING IS EQUIVOCALITY IN FORM. The poem concludes with another personiﬁcation, uo no me wa namida (‘there are tears in the eyes of ﬁsh’), which is a novel extension of the conventional metaphor, FISH ARE HUMAN. Although the personiﬁcation of ﬁsh appears frequently in folk tales and children’s stories, there are only a few linguistic manifestations of this metaphor in everyday idioms, e.g. kono sakana wa indo-yoo made tabi ni deru (‘this ﬁsh takes a trip to the Indian Ocean’). What ﬁsh do in the conventional FISH ARE HUMAN metaphor is prototypically a physical action of some sort rather than a mental reaction. In Example 3.5 (a), uo no me wa namida (‘there are tears in the eyes of ﬁsh’), is novel in that it extends the FISH ARE HUMAN metaphor to emotions; namely, the ﬁsh are crying in grief, shedding tears of sadness. It involves a very vivid and creative image of tears

Manifestation of Metaphor–Icon Links 79

Metaphorical Mapping

Iconic Mapping

Generic Space mouth

to wring

IMAGIC MAPPING

action of crying

NATURE IS HUMAN Personification

HUMAN in general human beings Basho friends

birds crying BIRD to cry

Particular PEOPLE

IMAGIC MAPPING

PEOPLE ARE NATURE Allegory crying of humans and non-humans

DIAGRAMMATIC MAPPING sadness felt by humans and non-humans Blend FORM

MEANING

EQUIVOCALITY IN MEANING IS EQUIVOCALITY IN FORM Grammatical Metaphor Figure 3.4 Metaphor–icon links in tori naki

in the human eyes mapped onto the eyes of the ﬁsh. An analysis similar to the BIRDS CRY metaphor may apply here, too. Fish are depicted in general with no speciﬁcation of number, size, shape, colour or name. As birds are a symbol for the sky, ﬁsh are a symbol for water in myth and folk belief. Hence, uo no me wa namida could imply metonymically that the water world shows sadness. On the other hand, an inverse allegorical reading, PEOPLE ARE FISH, is also possible. In Japanese, it is said that kare wa mizu o eta sakana no yoo da (‘he is like a ﬁsh swimming competently in the current’). In the poem, then, the personiﬁed ﬁsh which is (are) weeping in sadness could

80

Metaphor and Iconicity

signify either Basho, his friends, or both. In other words, the reading comes from the historical evidence that the agent of weeping could either be Basho or his friends, or both. 3.3.4 The autumn poem Overview of the autumn poem Basho wrote the following at the conclusion of his travel sketch: Rotsu had come as far as this port to welcome me back and accompanied me to the province of Mino. As I entered the manor of Ogaki, assisted by a horse, Sora came from Ise to join us. Etujin galloped his horse to do the same, and we all gathered in Joko’s house. Zensenshi, Keiko and his sons, and others who are close to me came to visit day and night and, as if meeting someone who had returned to life, expressed joy and consolation. Even before I was able to shake off the weariness of my travels, the sixth day of the ninth month came along, so I decided to offer prayers at the rebuilding of Ise Shrine and put myself on a boat again: Hamaguri no a clam GEN/SUB

futami ni wakare lid. ﬂesh into separate Futami (place name) to separate

yuku aki zo going autumn INT

‘A clam separates lid from ﬂesh as autumn departs’ (Matsuo 1996 [1694]: 133). Generally speaking, a similar analysis and argument hold for this poem. It is also a poem of farewell, and the season is the ‘passing autumn’. The poem’s more complex grammar and its sophisticated use of a pun betray Basho’s intent to be grammatically and lexically ambiguous. First, no in hamaguri no, can either be a genitive marker or a subject (nominative) marker. Second, futami functions as a multiple pun meaning ‘the clam-shell and clam-meat’, ‘to see the lid’, and the placename ‘Futami’. Third, wakareru is an intransitive verb of separation, meaning that something united or grouped as one becomes separate into two or more parts. Fourth, yuku in wakare yuku aki zo, can be grammatically connected either with wakare ‘to go separate ways’, or with aki ‘the autumn that is going away’. Without the contextual and cultural background, this poem sounds nonsensical or even funny. Why does the poet mention ‘a clam’? Why

Manifestation of Metaphor–Icon Links 81

Futami? What does ‘the clam-shell’ mean? The literal senses of the poem seem to lead to the two basic readings, both based on the futami pun. On the one hand, the poem literally describes natural events – that the autumn season is passing, the clam-shell and clam-meat separate. On the other hand, it depicts the human action of literally leaving for Futami (which is famous for good clams) or of leaving for the external Shrine of Ise (as metaphorised by the lid of a clam)15 in autumn. The blending model seems to explain how these two readings ‘blend’ in the poem. It is assumed that the two literal readings – the natural events of a separated clam-shell and the human action of leaving – represent two input spaces. The generic space has an event frame of separation common to these input spaces. The blend takes, from the input space of natural events, a strong image of the clam-shell and the clam-meat, once united but now in separate forms, and from the other input space of human action, an image of a traveller going his separate way to visit the renovated Ise Shrine. Each input space draws inferences derived from the local blends. Such inferences may include a strong sentiment of physical separation, the pain of a fatal wound from the ‘separation of the clam-shell and clammeat’. From ‘the passing autumn’, one may infer a sad feeling of parting of the traveller. Futami also has a double image of a place and a shrine. These inferences enrich the blend. In addition to the partial structure inherited from the inputs, the blend space has an emergent structure of its own – a fusion of death and life, or a death and rebirth. A vision of the blended double image of the death of the separated clam-shell and clam-meat is mapped onto the departing traveller, on the one hand, and the decline of the autumn season is mapped onto the renovation of the Ise Shrine, on the other. This subtle reading of contradictory images derived from the blend also echoes the passage preceding the haiku text: those ‘who are close to me came to visit day and night and, as if meeting someone who had returned to life, expressed joy and consolation’ (my italics) (Matsuo 1996 [1694]: 133). Local metaphors in the autumn poem The ﬁrst word, hamaguri (‘a clam’), together with the second word, futami (‘a lid and meat’) instantiates the traditional traits of haiku, that is, aisatsu (‘greeting’) and kokkei (‘wit’ or ‘a sense of humour’).16 The poetic form of haiku originated as hokku (ﬁrst haiku in the linked haiku texts) in the linked poetry called renku at the time of Basho. Hokku functioned as a greeting by a chief guest to the host of the poetry party

82

Metaphor and Iconicity

where linked poetry was a major entertainment. Hamaguri in this poem retains this trait of greeting, because both Ise where Basho composed this poem, and Futami, which is a town in Ise, are famous for good clams. To pay a compliment to something good relating to the host is a typical device of greeting. Hamaguri and futami also involve a blend of the HUMAN and the SHELLFISH inputs. The blended image of HUMAN and SHELLFISH can be witty or humorous, another trait of haiku. Basho’s choice of the word futami (‘a lid and meat’) presents a novel and witty image of separation. As mentioned earlier, two conceptual metaphors are interconnected here: NATURE IS HUMAN (or EVENTS ARE ACTIONS) and PEOPLE ARE NATURAL THINGS. First, the phrase, hamaguri no futami ni wakare, can imply that the clam-shell and the clam-meat are separated by the direct action of an agent (personiﬁcation). However, hamaguri (‘a clam’) is a uniﬁed whole made up of two elements, futa (‘a lid’) and mi (‘meat’), which are to separate by their own free will. The image is novel because the SHELLFISH domain is not generally exploited in the conventional NATURE IS HUMAN metaphor in Japanese. On the other hand, the inverse metaphor of PEOPLE ARE SHELLFISH is conventional in Japanese. The prototypical image-mapping of the PEOPLE ARE SHELLFISH metaphor is based on the image of the closed lids of shellﬁsh and the closed lips of humans, as in the idiomatic expression, kaigara no yoo ni kuchi o tozasu (‘to close the lips as if they were shells’). Thus, the image of the shells and the meat mapped onto human separation is a novel extension of the PEOPLE ARE SHELLFISH metaphor. The allegorical interpretation of the separation of the personiﬁed clam into its shell and meat is derived from the blend of two input spaces: the CLAM as a source and the HUMAN as a target. Basho’s parting from his friends and disciples with whom he was united at the end of the journey, the pragmatic knowledge provided by the context, is understood metaphorically as an image of separation of the shell from the meat of the clam. There are two other ways in which the image of the outside shell and the inside meat can be interpreted: functionally and ontologically. First, the shell functions as a protection for the inside meat, and so may be likened to the patrons and supporters who provided food, money and shelter to the poet on his travels. Second, the relationship of the shell and the meat involves ontological values17 held between the centre and the periphery, between the interior and the exterior.18 The clam, an animate creature in the centre, a metaphor for the poet, is of primary importance, whereas the shell, a protective cover, a metaphor for his friends (patrons and disciples) is of secondary importance, at least as

Manifestation of Metaphor–Icon Links 83

concerns the poet and his work. But the strength of the image of the separated clam-meat and shell comes from the factual knowledge that the clam no longer survives when its shell is separated from the meat. It is a forced separation against something united, in which some kind of force makes the shell open and separates the lid and the meat. In the blend, this image of fatal separation is mapped onto Basho’s departure on another journey. What is most notable in Basho’s choice of futami, however, is not simply this strong and novel image content, but also the multiple puns19 that epitomise and integrate the multiple interpretations of the poem, most notably the one about nature and the other about human beings. Each different reading of futami constitutes an input space in the blend, in which multiple meanings are integrated. As mentioned above, futami as a common compound noun of ‘a lid’ and ‘meat’ can be read allegorically as Basho’s friends (patrons) and Basho at the time of parting, based on the novel metaphor, SHELLFISH ARE PEOPLE, and the pragmatic knowledge of Basho’s farewell. There are two more readings of futami. Futami can be two words, a noun, futa (‘a lid’) and a verb, mi (‘to see’). Futa may then refer to an external Shrine of Ise, based on an imagemetaphor of exterior as discussed above and on the pragmatic knowledge of Basho’s visit to the Ise Shrine. This actually consists of two separate shrines: naiguu (‘the internal shrine’), and geguu (‘the external shrine’). The internal shrine houses the Sun Goddess, the most important goddess in Shintoism, while the external one houses Princess Toyuke, the goddess of food and industry. Here, too, the asymmetrical values in the internal/external relationship hold. Every 21 years, both shrines are renovated, with the ceremony for installing the new buildings taking place in September. Basho is recorded as having attended the ceremony for the external shrine, but missing the one for the internal shrine, which had been held three days earlier. Futami is also a proper noun, meaning Futami Cove, a bay near the Ise Shrine, and clams are one of the bay’s major products. Hence, in the blend of futami, several images and meanings derived from these readings are fused: the lid and meat of a clam, Basho and his friends, the internal and external shrines of Ise, and Futami Cove with its wealth of clams, as illustrated in Figure 3.5. The written form of futami also contributes to the multiple pun. As mentioned earlier, Japanese has three different writing systems: the use of Chinese logographs, kanji, and the two types of Japanese syllabary, hiragana and katakana. In revising this text, Basho changed futami in kanji, , which represent the place name exclusively because of the

84

Metaphor and Iconicity Metaphorical and Metonymic Mapping

Iconic Mapping

Metonymy (one meaning – Place Name) revision

(multiple meaning)

cove shrine

HUMAN in general human beings Basho friends

shellfish

Particular PEOPLE

SHELLFISH IS HUMAN Personification

PEOPLE ARE NATURE Allegory

DIAGRAMMATIC MAPPING Blend FORM

MEANING

EQUIVOCALITY IN MEANING IS EQUIVOCALITY IN FORM Grammatical Metaphor Figure 3.5 Metaphor–icon links in futami

speciﬁc logographs chosen, to futami in hiragana, , which just represent the moraic reading of the word (Matsuo 1957 [1694], 1985, Kaneko 1973). Figure 3.5 shows this revision in the box of ‘form’. Based on the grammatical metaphor EQUIVOCALITY IN MEANING IS EQUIVOCALITY IN FORM, the choice of hiragana writing for futami retains iconically the multiple meaning of futami in this poetic text. As futami has so many different readings, so the phrase, futami ni wakare, also displays a range of different interpretations. Wakare (‘to separate’) refers both to the action of human parting and the separation of the lid and meat of the clam. An image of the strong force of separation and even death of the clam, implied in the ‘lid and meat’ reading, is mapped onto the human farewell, the separation of Basho from his friends. The implication of death is also consonant with the season in which this haiku is set, that is the end of autumn. This is both the concluding phrase of the poem and of the whole of Basho’s travel sketch.

Manifestation of Metaphor–Icon Links 85

The last phrase of the poem, yuku aki (‘passing autumn’), is based on the conventional metaphors, TIME (SEASON) MOVES and TIME (and SEASON) IS A TRAVELLER. Also the metaphorical reading of AUTUMN as A TRAVELLER comes from the intra-textual knowledge that was explained earlier. In Japan, autumn is the season of two opposing aspects, bountiful harvest and inevitable decline as winter approaches. In the cognitive cultural model, however, the decline of progress and prosperity is stronger. For example, based on the LIFETIME IS A YEAR metaphor, there is a conventional conceptual metaphor, AUTUMN IS THE DECLINE OF LIFE, as seen in the following expressions: jinsei no aki (‘autumn in life’), and kare no jigyoo ni wa aki-kaze ga fui-te iru (‘there is an autumn wind blowing against his business’). The end of autumn, suggested by yuku aki (‘passing autumn’), is a period of cold winds, falling leaves and diminishing daylight hours, which gives a strong image-mapping between the NATURE and HUMAN domains in terms of decline. The verb yuku (‘to go’) in yuku aki zo implies, on the one hand, that autumn is drawing to an end, but perhaps on the other, departure for a new journey. The end of autumn is the end of an adventure and time for a much needed rest. It is also a metaphor of FAREWELL as a part of TRAVELLING, and signiﬁes allegorically that Basho is setting off on yet another journey as autumn departs. 3.3.5 Metaphor–icon links in macro-structure By way of a summary of the above discussion, it will be claimed that metaphor–icon links in the macro-structure of the travel sketch are isomorphic with those in the two haiku texts analysed. There seem to be two major conceptual metaphors contributing to the macrostructure as well as the micro-structure of the text: LIFE IS A JOURNEY and LIFE AND DEATH (REBIRTH) IS A CYCLE. LIFE IS A JOURNEY Basho is known as a poet who was on the road for much of his last ten years. Oku no Hosomichi was one of the several travel sketches that he wrote about his journeys. Basho, a ‘traveller poet’, stated his philosophy of life very clearly in the ﬁrst passage of Oku no Hosomichi: The months and days are wayfarers of a hundred generations, and the years that come and go are also travelers. Those who ﬂoat all their lives on a boat or reach their old age leading a horse by the bit make travel out of each day and inhabit travel. Many in the past also died while traveling. In which year it was I do not recall, but I, too, began

86

Metaphor and Iconicity

to be lured by the wind like a fragmentary cloud and have since been unable to resist wanderlust, roaming out to the seashores. (Matsuo 1996 [1694]: 41) Here, Basho exploits two conceptual metaphors, which are interrelated with each other. They are TIME IS A TRAVELLER and LIFE IS A JOURNEY. TIME IS A TRAVELLER

(a) The months and days are wayfarers of a hundred generations. (b) The years that come and go are also travellers. LIFE IS A JOURNEY

(c) Those who ﬂoat all their lives on a boat . . . (d) or reach their old age leading a horse by the bit . . . (e) [they] make travel out of each day and inhabit travel. On the one hand, the poet sees days, months and years, as travellers of eternity. On the other hand, one’s life from birth to death is a journey (boat journey or land journey on the horse). One travels each day; and travelling becomes one’s home. Compared to a ‘hundred generations’, one’s lifetime is very short. The double image of time merges here. Human life is a brief transient phase in eternity, which is travelled by time; nonetheless, human beings travel their lives each day, each month, each season and each year. As time passes by as a traveller, human life passes by. Furthermore, Basho’s philosophy of life as a journey can be placed in an even larger metaphorical framework in which every event and all things in the world are seen as a traveller. For the ﬁrst passage of Oku no Hosomichi alludes to the preface to Li Po’s (701–62) ‘Holding a Banquet in the Peach and Pear Garden on a Spring Night’. Li Po wrote: ‘Heaven and earth are the inn for all things, the light and shadow the traveler of a hundred generations. Accordingly, the ﬂoating life is just like a dream’ (quoted in Matsuo 1996[1694]: 40). It is noted that Basho took the phrase ‘the traveller of a hundred generations’ and rephrased ‘the light and shadow’ to ‘the months and days’. It was probably because Basho interpreted ‘the light and shadow’ as a metaphor for time.20 In addition, Basho sees life as maboroshi (‘an illusion’ or ‘a dream’) as he wrote that ‘[he] shed tears at a parting in this illusory world’ (Matsuo 1996[1694]: 43). This is a reﬂection from Li Po’s passage. The TIME IS A TRAVELLER metaphor is also detected in the use of a particular kanji, (‘to go’, ‘to pass’). As mentioned above, this logograph

Manifestation of Metaphor–Icon Links 87 tachikaeru revision yukikau Figure 3.6 Revision from tachikaeru to yukikau

is used in the two haiku texts which describe ‘passing spring’ and ‘passing autumn’. In the beginning passage of Oku no Hosomichi, Basho also used the same logograph when he said: ‘the years that come and go are also travelers (my italics)’ (Matsuo 1996[1694]: 41). The choice of this logograph to represent his philosophy of life was a deliberate one, because he revised the passage from tachikaeru (‘to start and return’) to yukikau (‘to come and go’). Notice the difference of logographs used as indicated in Figure 3.6 (Matsuo 1997 [1690]: 75). The repetitive use of the same logograph signals diagrammatically that the texts written by the use of this logograph represent similar (or repetitive) content. Therefore, it is clear that ‘passing spring’ and ‘passing autumn’ are a recapitulation of Basho’s philosophy as expressed metaphorically in the beginning passage of the sketch. In other words, the metaphor in the micro-structure, NATURAL SEASON as A TRAVELLER, is motivated by the prior textual metaphors in the macro-structure, TIME IS A TRAVELLER and LIFE IS A JOURNEY. In sum, the LIFE IS A JOURNEY metaphor is not only the metaphor which navigates the interpretation of the haiku texts which frame the travel sketch, but also the metaphor which represents the guiding philosophy of the entire sketch’s interpretation. The double function of this conceptual metaphor is reinforced by the iconic use of the logograph, . LIFE AND DEATH IS A CYCLE The fact that Basho sees life as an endless journey in the illusory world leads to the next conceptual metaphor, LIFE AND DEATH IS A CYCLE, which recurs in the travel sketch and which is epitomised in the two haiku texts as in Figure 3.7. The paralleled phrases, ‘passing spring’ and ‘passing autumn’, spelled in the identical logograph for ‘passing’, , function as an iconic frame of the whole sketch. The sameness of the logograph in the paralleled phrases at the beginning and the ending of the text mirrors the thematic unity. At the same time, this parallelism iconically hints at the grand cycle of being in the world, that is, a beginning is an ending; an ending is a beginning.

88

Metaphor and Iconicity

Yuku haru ya

tori naki uo no

me wa namida

Hamaguri no

futami ni wakare

yuku aki zo

Figure 3.7 Use of the logograph

in the paralleled texts

As mentioned earlier, spring is a season of birth and autumn is a season of decay and dying in the cognitive cultural model. It is therefore not far-fetched to read a cycle of life and death in the pairing of these two texts. In addition, the last haiku, the one that actually concludes the entire sketch, substantiates a fusion of death and life, or a death and rebirth. There is a double image of death and life. The death of the separated clam-shell and clam-meat is mapped onto the traveller who was departing for another new journey, while the dying autumn season is mapped onto the renovation of the Ise Shrine. The reading of death and birth in the paired texts in general, and in the last haiku in particular, is also consonant with the cyclic reading of the larger text: an end is a new beginning in life as in the journey. This is diagrammatically represented again by the use of the logograph, . The last word of the last poem, , iconically directs the reader back to the beginning, to the ﬁrst line on the initial page of the travel sketch that bears the same logograph, , as in Figure 3.8.

The last poem and the conclusion of the travel sketch

Hamaguri no

futami ni wakare

yuku ah zo

The first passage of the travel sketch

Tsukihi wa hyakudai no kakyaku ni shi-te, yukikau toshi mo mata tabibito nari. ‘the months and days are wayfarers of a hundred generations, and the years that come and go are also travelers’ (Matsuo 1996[1694]: 41) Figure 3.8 Use of the logograph travel sketch

in the beginning and the conclusion of the

Manifestation of Metaphor–Icon Links 89 The first passage of the travel sketch

The first poem of the journey

The last poem of the journey and the conclusion of the travel sketch

Figure 3.9 Use of the logograph

in the macro-structure of text

Consequently, the above discussion makes it clear that Basho’s choice of the logograph, , was a crucial one in both the micro-structure and the macro-structure. Figure 3.9 illustrates the use of in the macrostructure of text. The LIFE IS A JOURNEY and the LIFE AND DEATH IS A CYCLE metaphors navigate the iconic interpretation of the architecture of the entire sketch, and of the importance of the logograph, , in particular. The interplay of metaphor and iconicity shows up here. Without being guided by the conceptual metaphors detected from the thematic haiku and from the conventional metaphors in the culture, it is difﬁcult to read ‘in’ the diagrammatic iconicity of the structure of the text. Also, the iconic reading of the structure of the text reinforces the metaphorical interpretation of the blended text as a whole.

3.4 Concluding remarks This chapter has shown that the model of blending offers a useful analytical tool to explain the mechanism of iconic and metaphorical mapping in cognitive terms. An original contribution of this chapter is the demonstration of the effectiveness of this model for clarifying the complex interrelationship between metaphor and iconicity in the dynamic production of meaning in poetic texts. It has been demonstrated that close metaphor–icon links are manifest in two ways: that there are iconic (both imagic and diagrammatic) moments in metaphor; and that a form acquires an iconic meaning by metaphors. The analysis of the interplay between metaphor and iconicity has illustrated that the model of blending can specify which parts of the metaphorical process – whether the input, generic or blended spaces – relate to the diagrammatic iconic mapping of form and meaning.

90

Metaphor and Iconicity

Poems can be varied in their metaphor–icon structure. ‘The Easter Wings’ is dominantly iconic and has a much less rich use of metaphor, while ‘Love’s Philosophy’ is particularly rich and complex in its metaphor with respect to cognitive, image and grammatical aspects. These two poems are believed to be at relatively extreme ends of the spectrum and, to some extent, can be used as measuring standards for other poems to be analysed. The third text looked at two haiku in Oku no Hosomichi. The analysis illustrates that the interplay of metaphor and iconicity not only governs the micro-structure but also the macro-structure of a literary text. Some poems show metaphor–icon links in the visual and syntactic modalities. Others display such links in the auditory modality. Metaphor–icon links are not simple. In order to see the complexity and dynamism of the interplay between metaphor and iconicity, a detailed reading of the texts is needed in their various aspects – visual and auditory forms, structure of meaning, pragmatic and cultural contexts and background information. The following two chapters will attempt such detailed reading with special reference to the visual and the auditory manifestations of metaphor–icon links.

4 Vision as Meaning

Not a line is drawn without intention & that most discriminate & particular. As Poetry admits not a Letter that is Insigniﬁcant So Painting admits not a Grain of Sand or a Blade of Grass Insigniﬁcant, much less an Insigniﬁcant Blur or Mark. William Blake (1978 [1810])

4.1 Poetry and visual representation As William Blake (1978 [1810]: 1019), the poet and visual artist, wrote, both in painting and in poetry, the visual form can represent the conscious intention of the artist. The writing system, as I am going to argue in this study, not only refers to the conventions which record semantic and phonological content, but also covers a wider range of semiotic representations. These include the use of different kinds of typeface, spacing, indentation and layout, and they all have visual, graphic and conceptual effects. This wider sense of the writing system plays a crucial role in the way a poem is put on the page. Not only does it provide resources for the visual shape and structure of the poem, it can also inﬂuence the interpretation of the poem.1 This chapter will claim that visual iconicity, a link based on similarity between visible form and semantic content, is an important motivation in the generation and interpretation of poetic texts. The analytical potential of metaphor and iconicity will be examined in several illustrations of visual elements in Japanese and English poetry. Visual elements are roughly divided into two kinds: visible units such as letters, symbols and characters; and conﬁgurations of units, such as the line length, spacing, indentation, sectioning and general layout. To what extent poets exploit these visual options and how they actually 91

92

Metaphor and Iconicity

employ them in their own work depends on each poetic text. Also, the degree of awareness of the visual elements varies according to the poetic traditions and the status of written language in literary cultures. A comparative analysis of English and Japanese texts will present differing degrees of awareness due to differences between the status and nature of writing systems in these two languages and poetic traditions. The chapter, therefore, argues that regardless of the subtlety or prominence of iconic manifestations, once the iconic link is uncovered, it often directs or reinforces an interpretation of the text in that orientation. This is particularly noticeable in Japanese poetry where paying attention to the choice of character types and their layout in the text is a constant operation for both poets and readers. Furthermore, a comparison of English and Japanese poetic texts in visual terms renders greater awareness of, and more critical sensibility towards, the ‘neglected’ issue of the written language in the theory of poetry. Firstly, this chapter points out how English and Japanese differ from each other in their writing systems, and how their differences relate to the visual representation of English and Japanese poetry in general. Secondly, the chapter offers several illustrative analyses of the manifestations of the interplay between metaphor and iconicity in ‘visual poetry’ such as pattern poetry, concrete poetry and free verse. Lastly, more subtle manifestations of metaphor–icon links in conventional poetry are investigated with special reference to the relationship of orthographical revisions and the written language in haiku. 4.1.1 The writing systems of English and Japanese In order to achieve certain visual effects, poetry exploits various options in writing: for example, the size of letters, fonts, styles, the use of symbols and punctuation marks; the conﬁguration of the line, linelength, line-grouping, indentation, spacing and the layout on the page. Roughly speaking, these visual devices of poetry are divided into two kinds: units and conﬁgurations. The characteristics of units in writing is the focus of this section, whereas the following two sections (4.2 and 4.3) present a discussion of the conﬁgurations of these units, illustrated in the analyses of actual poetic texts. The visual units in writing used in English and Japanese are different in quality and quantity. English basically uses the phonographic alphabet system, whereas Japanese uses a mixed system of phonographic and logographic2 units.3 A closer look, however, reveals that both languages use both phonographic and logographic units, but to a differing degree.

Vision as Meaning 93

Modern English has two sets of phonographic units, i.e. 26 alphabetic symbols, in lower-case and upper-case letters, in addition to about a dozen logographic symbols such as @, #, &, and punctuation marks such as ‘, “, !, ?, ), ], <, :, ;, /, -, and a full stop. By contrast, modern Japanese uses two sets4 of phonographic units, i.e. 48 moraic characters, called hiragana (syllabary for words of Japanese origin and grammatical inﬂections and particles), katakana (syllabary for foreign words), which are extended to 71 by the use of diacritics,5 and a set of logographic units, i.e. several thousand Chinese characters, called kanji. Japanese also uses punctuation marks such as È˚ (‘single quotation’), (‘double quotation’), (‘full stop’), (‘comma’), and so forth. ° Not only because it is a mixed system of two different types of writing, phonographic and logographic, but also because it has an inventory of several thousand characters, the complexity of the Japanese writing system and the effect of this complexity on the visual associations of meaning deserve special attention in the examination of iconicity in poetry. As mentioned earlier, the Japanese language provides us with three character choices, i.e. kanji, hiragana and katakana as basic visual units. Each type of character has its own grammatical characteristics as shown in Table 4.1. Kanji, the Chinese logographic characters, are used mainly for words of Chinese origin as well as for the roots of such content words as nouns, verbs and adjectives of Japanese origin. Most kanji characters are built up from a limited number of basic constituents, called ‘radicals’. For example, the kanji for ‘rocks’ consists of two radicals, (‘mountain’) and (‘stone’). How radicals are put together is related to the meaning

Table 4.1 Characteristics of character types Character types

Characteristics

Kanji (Chinese logographic characters)

For words of Chinese origin and for the roots of such content words as nouns, verbs and adjectives of Japanese origin For words of Japanese origin for which there are no kanji, conjugated endings, conjunctions, particles, auxiliary verbs, and so on For words of foreign origin other than Chinese and for onomatopoeia

Hiragana (syllabary)

Katakana (syllabary)

94

Metaphor and Iconicity

and the pronunciation of the kanji. Hiragana characters, the Japanese moraic alphabet, are used mainly for words of Japanese origin, conjugated endings, conjunctions, particles and auxiliary verbs. Katakana characters, another moraic alphabet, are mainly used for words of foreign origin other than Chinese and for onomatopoeia. All three modes of representation – kanji, hiragana and katakana – are ordinarily combined in the written language. Each of these three types of character is different in nature and shape, and so each has different associations. In particular, kanji and hiragana, used to write Japanese words, have contrasting associations. Historically, kanji characters are the oldest and were used mainly by men. By the ninth century, a simpliﬁed form of kanji characters was devised as a set of syllabary, which became known as hiragana. Whereas men used both kanji and hiragana, women mainly used hiragana, which were called onna-de (‘female hand’). This historical fact has contributed to male/female associations of the character types, namely that kanji is masculine and hiragana feminine.6 Perhaps more signiﬁcant in terms of direct visual impact is the different shape of the characters. Kanji tend to have an angular form with more strokes and a distinctive shape, which gives rise to associations of sharpness, strength and tension, which many consider more masculine. On the other hand, hiragana characters tend to have a ﬂowing form with curved strokes, evoking associations of gentleness, softness and smoothness, which might be taken as more feminine. Hiragana and katakana are different in use and shape, too. Katakana characters have a more angular and distinctive shape than hiragana, and therefore give more masculine associations than feminine. In terms of visual effect, probably the difference between hiragana and katakana is comparable to the difference between lower-case and upper-case letters in the English alphabet. Like hiragana in Japanese, lower-case letters have curlier and smaller shapes than upper-case letters as in ‘a’, ‘e’, ‘h’, ‘m’, ‘n’ and ‘t’ vs. ‘A’, ‘E’, ‘H’, ‘M’, ‘N’ and ‘T’. The same word will look different when it is written in lower-case letters rather than upper-case letters, such as ‘time’ and ‘TIME’, or (in hiragana) and (in katakana). In addition to the psychological associations of their visual images, the two character types, logographic units (such as kanji) and phonographic units (such as hiragana, katakana and the English alphabet), have a different semantic integrity and expressiveness. Their major difference is that the former convey meaning itself in a presentational mode, whereas the latter do not as they are linear discursive signs.

Vision as Meaning 95

English and Japanese writing systems are markedly different, both in number and in conﬁguration of the units of representation. English uses considerably fewer phonographic letters in combination, while Japanese has at least three different systems of writing in operation on both the phonographic and logographic levels. Moreover, Japanese utilises a stock of several thousand characters. There is another important graphic difference relating to the conﬁguration of the units. In English, words and sentences are laid out with spaces in between, so that each word is processed graphically as a set unit. By contrast, Japanese words are written with no interval. However, those words written in kanji stand out as a set unit in a sentence, because of the density of their shape and the dominance of their visual–semantic link. The following examples, 4.1(a) and (b), illustrate the difference in the sentence, ‘Time is money’, in English and Japanese. Example 4.1 (a) Time is money. (b) The ﬁrst and the third characters, and , in Example 4.1(b) are kanji, representing ‘time’ and ‘money’ respectively. The other characters are hiragana, functioning as a topic marking particle, [wa], and a copula, [nari]. It can be said that in Japanese writing, words written in kanji are perceived as the ﬁgure and those in hiragana as the ground. Hence, the writer is always tasked with deciding whether to use kanji or hiragana to spell out the words, as the character choice gives different perceptions to the reader. A comparison of Example 4.1(b) (‘time’ and ‘money’ written in kanji) and 4.1(c) (all written in hiragana) demonstrates this point. Example 4.1 (b) (c) In 4.1(b), the two words, ‘time’ and ‘money’, are readily perceived and processed by the reader, whereas in 4.1(c), all the characters are perceived somewhat similarly so that the reader has to pay more attention to the phonological reading of the characters.7 In short, English and Japanese employ contrastive resources with respect to the basic units and conﬁgurations for visual effects. Bearing

96

Metaphor and Iconicity

these basic differences in mind, the following section experiments with possible variations of such visual resources in poetry. 4.1.2 Visual representation in poetry The haiku-like poem entitled, ‘In a Station of the Metro’, by Ezra Pound (1928 [1916]: 53) is used here to examine how different visual representations, created by the manipulation of letters, sizes, fonts and layouts, can affect impressions of the poem: Example 4.2

Example 4.3

IN A STATION OF THE METRO

THE APPARITION OF THESE FACES IN THE CROWD PETALS ON A WET, BLACK BOUGH. Example 4.4

IN A STATION OF THE METRO

The apparition of these faces in the crowd; Petals on a wet, black bough. Example 4.5

IN A STATION OF THE METRO The apparition of these faces in the crowd; Petals on a wet, black bough.

Vision as Meaning 97

Example 4.6

IN A STATION OF THE METRO The apparition of these faces in the crowd; Petals on a wet, black bough. Example 4.7

IN A STATION OF THE METRO The apparition of these faces in the crowd; Petals on a wet, black bough. Example 4.2 is a reproduction of the extract from the Faber edition of Selected Poems of Ezra Pound (1928 [1916]: 53). Example 4.3 uses uppercase letters. If this passage were pasted on a bulletin board, it might give an impression of a notice rather than a poem because of the emphatic visual tone of the capital letters. Example 4.4, written in 11-point Courier font, has a likely association with a typewritten script. Example 4.5 uses

98

Metaphor and Iconicity

a much larger size of Helvetica font in boldface, which may convey a sense of graphic sophistication, or of commercial advertisement. Examples 4.2–4.5, in spite of differences in font size and style, are faithful to the original line conﬁguration. Examples 4.6 and 4.7, on the other hand, experiment with different layouts of lines. Example 4.6 is a 3-line conﬁguration of 5- and 7-syllable lines, an attempt to make the poem resemble the appearance of Japanese haiku poems on the page. Together with the Monaco font in italics, which suggests a non-ordinary print form, this example may convey something distinct and foreign. Example 4.7 is even more radical in the way each line is laid out on the page. The vertical shape of the poem tries to capture the shape of a black bough. The scattered words represent petals on the wet bough, which are a metaphor for the apparition of faces in a crowd.8 By comparison, a haiku can be presented visually in three different writing systems by the use of hiragana (syllabary for words of Japanese origin and grammatical inﬂections and particles), katakana (syllabary for foreign words) and kanji (Chinese characters), written either horizontally or vertically, in addition to a variety of font, size, style, layout, etc. For example, a haiku, ‘shizukasa ya, iwa ni shimiiru, semi no koe’ (‘Quietness: seeping into the rocks, the cicada’s voice’) (Matsuo 1996 [1694]: 95) can be represented in many different ways: Example 4.8 Example 4.9 Example 4.10 Examples 4.8–4.10 horizontally reproduce the original text in three different fonts. Example 4.8 is an approximate reproduction of the standard Japanese printed texts, written in Mincho font in plain style. The use of a Gothic font in boldface in Example 4.9 conveys a somewhat contemporary atmosphere because this font often appears in modern advertisements of industrial design. In contrast, Example 4.10 simulates the handwritten version done by calligraphy, and gives a sense of antique authenticity. It is clear that the visual impression is affected by the use of different sizes, fonts, etc. in both English and Japanese. Example 4.11 Example 4.12 Example 4.13 Example 4.14

Vision as Meaning 99

Examples 4.11–4.14 display the possibilities unique to Japanese. These examples use the same font in the same size and style; however, they are written in different character types. Example 4.11 is written entirely in hiragana, while Example 4.12 is written entirely in katakana. Example 4.13 uses hiragana for the ﬁrst word, shizukasa (‘quietness’), instead of kanji in the original version, while Example 4.14 uses katakana for the second last word, semi (‘cicada’), instead of kanji, , in the original. As explained in the previous section, hiragana notation is primarily perceived as ground against ﬁgure. In Example 4.11, there is no kanji to be perceived as ﬁgure, and hence, the sequence in hiragana is hard to read at ﬁrst glance. Likewise, in Example 4.12, katakana notation is also perceived as ground. In Example 4.13, as shizukasa (‘quietness’) is written in hiragana, it gives a sense of gentleness or softness, whereas semi (‘cicada’) written in katakana in Example 4.14 may give emphasis or a foreign connotation to the word, as katakana is primarily used for the words of foreign origin. Moreover, the ﬁrst word in Example 4.14 is written in a different kanji, , than the original, , in order to show that the same idea (‘quietness’ in this case) can be represented in several different kanji characters. Each character has a different etymology and thereby subtly differing nuances. For example, the kanji, , in Examples 4.8–4.10 etymologically means to put the bar on the gate. The idea is that serenity comes when the place is closed. On the other hand, the etymology of the kanji, , in Example 4.14 means to stop the ﬁght. The idea is a peaceful silence. The choice of kanji is dependent on the content of the text in which it is used, the overall context, style of writing and the preference of the writer (for further discussion, see Section 4.3.2; also Chapters 3 and 7). It should also be noted here that a vertical layout is more common in Japanese writing. A text is written from the top of the page downward and from right to left. Literary works in particular are written vertically. Example 4.15 shows the vertical representation of the text of the poem in Example 4.8. Example 4.15

100 Metaphor and Iconicity

It becomes clear from the above observation that in Japanese writing, the choice of character type or of a certain kanji character is a constant operation for any Japanese writer. Poets, however, are highly conscious of this selection process; indeed they deliberately exploit a wide range of potential combinations. Thus, the complexity of the Japanese writing system affords a dynamism and freedom to the visual representation of Japanese poetry. The following analysis shows how poetic form visually embodies meaning in overt manifestations of iconicity in ‘visual poetry’ and more subtle manifestations in traditional poetry in English and Japanese.

4.2 Visual poetry The phrase ‘vision as meaning’ in poetry is most likely to cue associations of ‘visual poetry’. In visual poetry, how the poems are laid out on the page, i.e. their shape and structure, is a primary means of conveying formal effects concurrently with the lexical content of the poems. Visual poetry embraces several kinds of poems such as pattern poetry, concrete poetry and free verse. The poems mark different poetic movements in history as well as differing degrees of dependence on the concrete visual shape of poetic texts.9 Pattern poetry is straightforwardly mimetic of the shape of objects that the poems describe, while concrete poetry and some attempts in the tradition of free verse display geometric ﬁgures or structural patterns which render the theme or the subject of the poem only indirectly in an abstract analogy. The purpose of this section is to illustrate and analyse the function of visible representations in covert manifestations of metaphor–icon links in a few sample visual poems in English and Japanese. 4.2.1 Imagic manifestations As stated before, poems show up iconicity and metaphor in a complex way so that it is difﬁcult to single out ‘imagic’ iconicity alone. However, it can safely be said that there are visual poems, particularly pattern poems, in which imagic iconicity is a dominant factor, and that it is such imagic iconicity that leads to or reinforces the semantic content of the poems.10 ‘Swan and Shadow’ Example 4.16 (Hollander 1991[1969]: 35)

Vision as Meaning 101

John Hollander’s ‘Swan and Shadow’ is perhaps one of the best-known modern efforts to entertain the eyes, along with the same tradition of pattern poetry as seen in ‘Easter Wings’ by George Herbert (1880 [1633]: 34–5) (for a detailed analysis of ‘Easter Wings’, see Chapter 3). The shape and layout, and the title of the poem are so compelling that it is almost impossible to interpret this poem irrespectively of ‘swan and shadow’. Even when one reads the text at ﬁrst, one’s attention goes to the choice of words which is constrained by the form rather than to the lexical contents. For example, the neck requires short words, and the reﬂection requires careful choice to ensure that it is an exact replica of the object. The question/answer format of the upper neck/wing is repeated in the reﬂection, even including the use of capital letters. It is not easy to read this poem naturally, because its shape is so striking. Therefore, if the text is written as below in a prosaic version, the original impact of the poem as an icon is completely lost.

102 Metaphor and Iconicity

Example 4.17 Dusk Above the water hang the loud ﬂies Here O so gray then What A pale signal will appear When Soon before its shadow fades Where Here in this pool of opened eye In us No Upon us As at the very edges of where we take shape in the dark air this object bares its image awakening ripples of recognition that will brush darkness up into light even after this bird this hour both drift by atop the perfect sad instant now already passing out of sight toward yet-untroubled reﬂection this image bears its object darkening into memorial shades Scattered bits of light No of water Or something across water Breaking up No Being regathered soon Yet by then a swan will have gone Yes out of mind into what vast pale hush of a place past sudden dark as if a swan sang Semantically, the poem points to the ﬂeeting existence of a swan and its shadow on the surface of water. The shape of this poem seems to capture this essential moment of the poem’s content. In this sense, the poem’s shape is a strong imagic manifestation of its content. However, the experience of the poem goes even deeper if we slowly turn our attention to the meaning of the poem. This is an intriguing poem in which the narrator is in fact a personiﬁed tiny droplet of water on the surface of a pond. It is dusk, silent and peaceful apart from the evening buzzing of ﬂies above the surface, until the grey monotony is broken by the shadow/reﬂection of a passing swan and the accompanying ripples.

Vision as Meaning 103

The drop as an individual is intimately joined with the millions of other droplets in the pool that is their world, and that is also an eye opening onto the dark air above – but an eye of what we are not quite sure. At the same time, cleverly evoking the effect of ripples on a reﬂection, the poem suggests that the passing swan’s reﬂection, perhaps some magical or unfathomable event, brings lightness, happiness and interest into the otherwise melancholic and dull present. But even if the swan was recognised for what it is, it is soon forgotten as the perfect calm surface returns and the short hour of dusk drifts on into night. At night, the eye is closed and the droplets appear to have little conception of this mysterious calm. This simple scene reﬂects the transient beauty or ﬂeeting truth of nature observed from the perspective of a surface drop of water in personiﬁcation. Suppose this poem as a global metaphor has two input spaces: the swan and its reﬂection in the source input space, and the ﬂeeting moment of perfect sadness felt by the personiﬁed droplet in the target as illustrated in Figure 4.1. The shape of the swan, of its shadow and of the surface of the water is mapped onto the source input space by imagic iconicity. In the blended space, there is an emergent image of the sudden grasp of this moment of perfect sadness or the ‘nowness’ of the present instant, in addition to the passing swan and its reﬂection and the ﬂeeting motion of time. There is a contradictory force mediating between moving objects (the swan, its shadow, and time) and the cessation of such movement (the moment of sadness). This grasp of the frozen moment of sadness then navigates the reading of the visual representation. Mediated by the grammatical metaphor, A STATE IS NON-LINEARITY, or A PICTURE FREEZES MOVEMENT (OF TIME AND OF OBJECTS), the blended space is mapped diagrammatically onto the visual form of the poem. The picture of the swan and the shadow captures the ‘perfect sad instant now’ on paper as a movement capturing the ﬂow of time, something more difﬁcult to achieve by the conventional poetic form which, by the nature of language itself, is linear and discursive. Another diagrammatic mapping occurs between the form and the blend. The central line in the form, ‘even after this bird this hour both drift by atop the perfect sad instant now’, is the punchline of the poem. Namely, it is the line which materialises the blend – the bird, the time, the perfect sadness, the surface of the water. Thus, one can detect a diagrammatic manifestation of the grammatical metaphor, IMPORTANT IS CENTRE. Likewise, the longest sentence of the poem starts in the middle of the body of the swan, continues to the surface of a pool, and forms

104 Metaphor and Iconicity

Iconic Mapping

Metaphorical Mapping

Target

Source Visual Shape Swan Surface Shadow

IMAGIC MAPPING

DIAGRAMMATIC MAPPING

Blend FORM

MEANING

A STATE IS NON-LINEARITY IMPORTANT IS CENTRE Grammatical Metaphors Figure 4.1 Metaphor–icon links in ‘Swan and Shadow’

a part of the swan’s shadow. The extraordinary length of this sentence makes one wonder if the poet deliberately prolongs the sentence to mean graphically and analogically that there is no clear separation between the swan, the surface and the shadow; that there is a continuum of object and image, light and darkness, present and past. In a way, it is perhaps ‘the perfect sad instant now’ that is just happening momentarily. It is clear that, in a pattern poem of a very distinct shape such as Example 4.16, the interplay of metaphor and iconicity is manifested most predominantly in imagic mapping between form and meaning. It is the form that navigates the interpretation in the direction of the visual shape. Nonetheless, a closer reading of the poem reveals that metaphor–icon links are of a more complex nature even in a seemingly simple example of pattern poetry.

Vision as Meaning 105

‘40–Love’ The next example is somewhat more obscure in shape; but it neatly displays the interplay between metaphor and iconicity. Example 4.18 (McGough 1989 [1971]: 84–5) 4o–

Love

middle

aged

couple

playing

ten

nis

when

the

game

ends

and

they

go

home

the

net

will

still

be tween

be them

Most readers are likely to ask why it is laid out in the way it is. The shape and layout alone may suggest some ‘ﬁgurative’ interpretation of the visual text; but, what kind of ‘ﬁgurative’ meaning can be drawn out depends predominantly on the content (both the title and the text) of the poem. So, the next step is to read the words. The poem reads as in Example 4.19. Example 4.19 40–Love middle-aged couple play tennis when the game ends and they go home the net will still be between them Although the original strength of the poem’s iconicity disappears in the prosaic presentation in Example 4.19, the reader is drawn to its semantics. The poem is about tennis. It is about a middle-aged couple. It is about the net between the two players, between the couple. The middle-aged couple is compared to tennis players. They are not doubles

106 Metaphor and Iconicity

partners. They are playing against each other. The score of the game is 40–love. One is about to win the game. At the same time, it is a pun: love at the age of forty, the relationship of a middle-aged couple. The game of tennis is over. But their real game (of love) is not over, because there is still the net between them. The relationship of this couple is further compared to the net dividing the tennis court. This text can be seen as a metaphor in which a love relationship is understood in terms of a game of tennis. Images of tennis players, the game they are playing, and the net dividing them are mapped onto the images of the couple, their love relationship. The correspondence of these images constitutes a schematic or diagrammatic mapping of the metaphorical process, which in turn enhances the blend, a fused structure of the one-sided, tired relationship of a middleaged couple. With this metaphorical reading, the visual conﬁguration of the text shows up its iconic meaning. The facing two pages are the tennis court, and the binding line of the pages is the net. Words in the text are either monosyllabic or disyllabic. Besides, some disyllabic words are cut into two monosyllabic units and displayed across the binding line. These small lexical elements are repeated in pairs 12 times in a row over the two pages as if they were the monotonous movement of a tennis ball across the net. It is important to note that this imagic iconicity is guided by the metaphorical reading of the poem, rather than that the image of the poem on the page guides the interpretation of the poem as it has been seen in the previous examples of pattern poetry (‘Easter Wings’ and ‘Swan and Shadow’). Hence, as indicated in Figure 4.2, the interplay of metaphor and iconicity in ‘40–Love’, consists of three major processes: the metaphorical mapping between the source and the target input spaces; the diagrammatic mapping of the blended space (metaphorical reading of the poem) onto the visual form of the poem; and the imagic mapping of the visual text onto the source input space of a tennis court and tennis game. In short, the metaphor of tennis navigates the iconic interpretation of the structure of the text.

4.2.2 Diagrammatic manifestations The next group of examples of visual poetry is classiﬁed as ‘concrete poetry’, which displays a more abstract structure than pattern poetry. Brogan (1994: 43) deﬁnes concrete poetry as

Vision as Meaning 107

Iconic Mapping

Metaphorical Mapping

Source Visual Shape

(3)

(1)

tennis net

tennis IMAGIC MAPPING ball game net

couple

game

DIAGRAMMATIC MAPPING (2)

barrier

Target relationship love

middle-aged one-sided Blend

FORM

MEANING

Figure 4.2 Metaphor–icon links in ‘40–Love’

visual poetry . . . , in which (1) each work deﬁnes its own form and is visually and, if possible, structurally original or even unique; (2) the piece is without any major allusion to any previously existing poem; and (3) the visual shape is wherever possible abstract, the words or letters within it behaving as ideograms. The poetic challenge of concrete poetry is to mediate between visual art and verbal art. In fact, the dominant proportion of concrete poems in the major anthologies (Williams 1967, Wildman 1969, Solt 1970, among others) shows an orientation towards visual abstract art. The individual letters or words give a poem its basic textural material, just as pigments – their colour and material – do for painting. The abstract conﬁgurations of visible linguistic elements provide an overall structure as the arrangements of colour, line and shape give paintings their perspective and composition, which match with the theme of the work. Concrete poetry is thus ‘concrete’ in the sense of its visibility; but the relation of its form with its content is of an abstract nature. Due to this abstract and structural orientation, diagrammatic iconicity predominates in this type of poetry. What is common in the manifestation of metaphor–icon links in concrete poetry is that the abstract visual shape and conﬁguration of

108 Metaphor and Iconicity

Metaphorical Mapping

Iconic Mapping

Source

(2)

Target

Visual Shape

DIAGRAMMATIC MAPPING (1)

Blend FORM

MEANING

Grammatical Metaphors Figure 4.3 Metaphor–icon links in concrete poetry

elements provide the main force for the interpretation of the text. Although the lexical and sentential meanings constitute the source to be mapped onto the target in metaphorical mapping, this metaphorical mapping seems to be a result of the diagrammatic mapping from the form onto the target in concrete poetry. It can be said that unlike the other forms of visual poetic texts analysed above, the diagrammatic conﬁguration of concrete poetry is initially mapped onto the target rather than the source, as illustrated in Figure 4.3, by a variety of grammatical metaphors such as MORE MEANING IS MORE FORM, DIFFERENCE (SIMILARITY) IN MEANING IS DIFFERENCE (SIMILARITY) IN FORM, and so on. Grammatical metaphors navigating the interpretation of the abstract form differ according to each poetic text; but the basic mapping structure in terms of diagrammatic iconicity is the same in the following examples analysed.

Vision as Meaning 109

‘l(a’ The following example by E. E. Cummings (1972 [1958]: 673) presents a radical poetic challenge, experimenting with a novel use of letters, symbols and words, and succeeds in evoking powerful images and readings. Example 4.20 (Cummings 1972 [1958]: 673) l(a le af fa ll s) one l iness

This poem is predominantly visual and structural, and even enigmatic at ﬁrst glance. There is only one line, Line 7, which looks grammatically English by containing the word ‘one’ as such; but later the reader will discover that it is a disguise, as it is only a part of another word ‘loneliness’. The poem has two parts: a word ‘loneliness’, spread over four lines, Lines 1, 7, 8 and 9; and a sentence, ‘a leaf falls’, put in parentheses, laid out vertically. These two parts are juxtaposed in a novel fashion, and suggest a metaphorical reading of the poem. Namely, there are two input spaces, NATURE and MIND, as epitomised in the images of a falling leaf and a feeling of loneliness. These spaces are ‘blended’11 in the juxtaposition in a way that a falling leaf is mapped onto loneliness. Just one leaf falls, a metaphor of solitude. The only ‘readable’ English line, ‘one’ in Line 7, is consonant with this metaphorical reading. Notice also that the word ‘one’ is a part of the word ‘loneliness’, as if it were one of the radicals constituting a Chinese character of ‘loneliness’.12 It is a leaf which is falling, a metaphor of decline, which is also visually presented by the vertical layout of letters and symbols.

110 Metaphor and Iconicity

This graphic play with letters and symbols is particularly characteristic of this poem. There are several manifestations of imagic iconicity in the poem: 1. The shape of the poem forms a single vertical line, which resembles the fall of a leaf, on the one hand; and on the other hand, it resembles a Roman numeral ‘one’, which conceptually represents ‘singularity’, which in turn leads to ‘loneliness’; 2. The use of l’s and f’s (vertically longer letters) reinforces the verticality of the fall; 3. Five sections of the poem, consisting of one or three lines alternating, may suggest a leaf ﬂoating down to the ground gradually; 4. The ﬂat shape of the last section ‘iness’ may suggest a leaf on the ground, the ﬁnal stage of the fall; 5. The shape of the parentheses represents the shape of a leaf; and 6. The seventh line is the only line that looks English, although it is a part of the word ‘loneliness’. This line could be intended to give the double meaning of ‘singularity’ and ‘loneliness’. The following line, Line 8, resembles an Arabic numeral ‘one’ rather than an alphabetic ‘l’ if we see it graphically, which again conceptually represents ‘singularity’ that leads to ‘loneliness’. On the more abstract level, the poem can be further read as a structured diagrammatic icon at least in the following three aspects: 1. The use of parentheses graphically represents the uniﬁed and yet suspended image of a falling leaf and loneliness; 2. At the same time, the use of parentheses forces the reader to go back to the ﬁrst line after the sixth line. The reader has to discover a connection between the ﬁrst ‘l’ and the ‘one’ in Line 7, which follows the parenthesis. The two themes of the poem are mediated by this discovery; and 3. The central section, which consists of two l’s, graphically unites the upper part and the lower part of the poem, a falling leaf and loneliness, as if a mathematical equation symbol ‘=’ were drawn vertically. As the analysis above demonstrates, the poem visually manifests imagic and diagrammatic iconicity. Yet it is also a creative attempt to indicate a visual metaphor, which embodies the juxtaposition of two

Vision as Meaning 111

images, or two conceptual domains, put in tension, in order to offer a new way of seeing the world. It is interesting to see that there are a number of parallels to English in the diagrammatic manifestations of Japanese visual poetry. The graphics of the Japanese examples dominate the reading experience because of the use of different character types, particularly the logographic kanji characters. Nevertheless, the basic manifestations of iconicity are the same, as seen in the following analyses. ‘Ame (rain)’ Example 4.21 (Niikuni 1985 [1966]: 21)

This poem, entitled ‘Ame (rain)’, is probably one of the best-known examples of Japanese concrete poetry. It is simple, and yet very powerful in its composition. There are 19 repetitions per line and 24 lines of the ‘four dots’, which are part of the logograph, (‘rain’), in the bottom line. The abundance of the repetition of the dots with the white background appears to signify the endless movement of rain. The poem is

112 Metaphor and Iconicity

a mimetic image of falling raindrops. The recognition of the kanji, , as a moment of change into non-repetition precedes the recognition of the repetitive background against which it is foregrounded. The process of the eye movement, from the bottom to the top, is also iconic to our actual recognition of rain. What is intriguing to me in this particular example, however, is its diagrammatic iconicity, namely, that the poem is a challenge to our recognition of the kanji, and as such is an example of the technique of defamiliarisation. For this poem shows us in a totally estranged context – a mass of repetitive ‘four dots’. This estranged presentation converts the transparency of into an opaque entity, which in turn serves as an opportunity to revive its etymological reading, that is raindrops and a heaven where raindrops originate. In the same manner, the very presence of in the poem gives meaning to the repetition of ‘four dots’, which would otherwise be interpreted as graphical dots and not as raindrops. Therefore, the poem displays itself as a diagram, too, iconically representing a change from an ordinary to an extraordinary recognition of , which is foregrounded against the extraordinary context – the repetition of ‘its parts’. It seems obvious that the dominant repetition of ‘four dots’ displays an iconic meaning, signifying the endless, continuous abundance of raindrops. Likewise, the change from the repetition of ‘four dots’ to non-repetition iconically signals the foregrounding or the distinctiveness of the kanji . In this example, the effect of this foregrounding is multiplied because the repetitive unit is a part of a character and the non-repetitive unit is a character. In other words, the difference in visual density and semantic weight between the ‘four dots’ and the kanji iconically emphasises the distinctiveness of the latter, which etymologically means a heaven and raindrops.

‘Fuukei (scenery)’ The next example (Yamamura 1989 [1915]: 89–90) illustrates a repetition of larger units in similar manifestation types, but with different semantic effects. Example 4.22 (Yamamura 1989 [1915]: 89–90)

Vision as Meaning 113

‘Fuukei (Scenery)’ Transcription

Word-for-word translation

I 1 2 3 4 5 6 7 8

Ichimen no nanohana Ichimen no nanohan Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Kasukanaru mugibue Ichimen no nanohana

1 2 3 4 5 6 7 8

A ﬁeld of turnip.ﬂowers13 A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers Faint straw.ﬂute A ﬁeld of turnip.ﬂowers

1 2 3 4 5 6 7 8

Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Hibari no oshaberi Ichimen no nanohana

1 2 3 4 5 6 7 8

A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers Lark of chattering A ﬁeld of turnip.ﬂowers

1 2 3 4 5 6 7 8

Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Ichimen no nanohana Yameru wa hiru no tsuki Ichimen no nanohana

1 2 3 4 5 6 7 8

A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers A ﬁeld of turnip.ﬂowers Pallid TOP day of moon A ﬁeld of turnip.ﬂowers

II

III

114 Metaphor and Iconicity

This visual poem describes a scene of countless nanohana (‘turnip ﬂowers’) blooming in a ﬁeld. A massive yellow stretch of innumerable tiny ﬂowers spreads before our eyes. The faint sound of a straw ﬂute is heard in our ears. The larks are chattering over our head. When we turn our eyes up to the sky, we see the pale ﬁgure of the moon hanging in the sunny sky. The repetitions characteristic of this poem use imagic and diagrammatic iconicity. Firstly, the poem consists of three stanzas of eight lines. Each stanza has seven repetitions of ichimen no nanohana (‘a ﬁeld of turnip ﬂowers’) with a single variant on the seventh line. Each line has nine morae, written in nine hiragana. Hence, iconically the repetition of the same character type gives a monotonous, or a levelling graphic effect (diagrammatic manifestation); the repetition of the same number of lines with the same number of characters of the same type gives each stanza an exact rectangular shape, which resembles a patch of turnip ﬂowers (imagic manifestation). Secondly, the use of moraic hiragana instead of logographic kanji is particularly important to this poem, because it not only gives an exact rectangular shape to each stanza but also deceives our eyes into overlooking the difference of the seventh line from the repetitious lines due to its similarity of shape and number. The content of the seventh line in each stanza – the faint sounds of a straw ﬂute, the chattering of larks and a pale moon in the day sky – is what we might assume to be buried in the strong and massive yellowness of the ﬂowers. The way the poem displays the seventh line in a kind of disguise is diagrammatically iconic to its meaning. Thirdly, the overwhelming sequence repetition of ichimen no nanohana emphatically suggests the dominance of the turnip ﬂowers in the scenery. At the same time, the change from repetition to non-repetition occurs on the seventh line. Once we have recognised the seventh line as distinct, it is foregrounded and catches our attention. Yet, because it is less dominant than the background of sequence repetition, it can slip from our attention in the same way as do its semantic referents cognitively. Fourthly, the placement of the distinctive line seventh among the eight lines of stanzaic composition also has an iconic meaning. It is necessary to have six lines of repetition to create ‘mass’ in order to emphasise the continuation of ‘the ﬁeld of turnip ﬂowers’. It could not occur in the last line of the stanza, because it would frame the stanza and would stand out too much to be disguised. The framing lines of each stanza – the ﬁrst and the last – are repeated, and provide framing to the

Vision as Meaning 115

stanza. The framing lines also have a cyclic effect, which suggests iconically an indeﬁnite continuation of the framing elements – namely, the ﬁelds of turnip ﬂowers. Lastly, the fact that each stanza has a variant seventh line is another manifestation of repetition – repetition at similar intervals. It shows that the stanzaic structures are equated to each other, having a regular recurrence of a similar theme by the repetition of a variant line in the same stanzaic position. Therefore, the following becomes clear: that repetition in quantity conveys dominance, continuation, monotony, a levelling effect, background and the like as iconic meanings; that a change to non-repetition conveys foregrounding; that repetition at the beginning and the end gives a framing or a cyclic effect; and that repetition at similar intervals conveys regularity and/or unity. 4.2.3 Remarks on comparison There are two major conclusions to be drawn from the comparison of iconic manifestations in ‘visual poetry’ in English and Japanese. Beginning with what they have in common, since visual poetry is, by deﬁnition, characterised by its orientation towards the ‘visible’ shape, conﬁguration and structure in relation to the semantic content of the texts, those poems classiﬁed in this category of poetry in English and Japanese both visually display imagic and diagrammatic iconic manifestations. In general, pattern poetry is dominantly imagic, whereas concrete poetry and free verse manifest more diagrammatic iconicity in the structural analogy between the visual form and the semantic content. If we look at the poems as global metaphors, iconicity interacts with the metaphorical content of the poem. The complexity of the interplay between metaphor and iconicity varies from poem to poem. However simple it may look, the poetic text is a complex piece of artwork, which condenses multiple manifestations of iconicity in most intricate and exquisite ways. In terms of the conﬁgurations of the units, the poetic texts looked at in both languages show a remarkable similarity. It should be noted, however, that such a striking similarity is partly due to modern crosscultural inﬂuences. As mentioned above, pattern poetry in Japanese is indebted to European pattern poetry. The concrete poems in English are in the Imagist tradition of japonaiserie. Nonetheless, the point made with regard to visual iconic manifestations is that the visual poets exploit in their linguistic and semiotic resources every conceivable combination and repetition, which not only reveal novel and unique ways

116 Metaphor and Iconicity

of understanding the objects described in the poetic texts, but also create new ways of appreciating the poetry itself. Visual poets cross the boundaries between visual art and verbal art. Visual poetry is thus motivated towards presenting a new way of ‘looking at’ poetry. That which distinguishes English from Japanese visual poetry depends largely on the nature of the written scripts each language employs. As mentioned in Section 4.1, the English writing system is dominantly phonographic, whereas the Japanese system is a mixture of logographic and phonographic units. The logographic orientation of the writing system itself seems to generate a stronger visual iconic motivation, which materialises in Japanese poetry in both imagic and diagrammatic forms. Japanese poets, not only visual poets but poets in general, are consciously aware of the choice of characters in the writing of poetry;14 unless they compose their poems entirely for oral presentation. Rich resources are available, particularly in the inventory of kanji characters. In Japanese concrete poetry (such as Example 4.21), kanji are a dominant element. Precisely because kanji are logographic, the main interest of visual poetry in Japanese rests more on the conﬁgurations. Such poets do not have to produce a new unit which functions as another ideogram. They can use what they already have. By contrast, English visual poets, particularly concrete poets, try to create their own units to be conﬁgured as if they were ideograms. Cummings is one of those poets who challenge their own writing systems to create radical and innovative ideograms of their own. Cummings, for example, dissects English words into smaller units, in either a letter or a group of letters, and arranges them in an iconic order so that some of them visually represent certain ideas. Because of the phonographic nature of the writing system, English visual poets have to make a more conscious effort to create visible units, compared to their Japanese counterparts.

4.3 Subtle manifestations of visual iconicity in poetry Visual poetry is the extreme case but not the only case in which the visual shape or visual conﬁguration of poetry conveys the meaning iconically. Even traditional poetry written in closed form (cf. Kennedy 1986: 157–67) uses graphic space in certain ways to produce abstract effects when printed on the page. Generally, the written form of poetry follows certain conventions. Such conventions include how the line is marked, the section is divided, the indentation is made, and so forth. For example, in traditional English poetry with rhyme and metre, a new

Vision as Meaning 117

line of a poem often starts with a capitalised word. A section is marked by a blank line. These conventions are not normally regarded as iconic, because they are taken for granted as conventions. However, a closer examination of the relationship of such formal appearances with the semantic content often leads to the realisation of some abstract diagrammatic correspondences between the form and meaning in grammatical metaphors. In this section, visual diagrammatic manifestations in traditional poetry in English and Japanese will be looked at with special reference to Japanese haiku. The peculiarity of the writing system of Japanese makes both writers and readers more aware of the appearance of the text, and hence, there are more visual iconic motivations both in the generation and the interpretation of the conventional Japanese poetic text than there are in English.15 4.3.1 General diagrammatic features Diagrammatic icon by grammatical metaphors There are general diagrammatic features of a visual nature embedded in the conventions of poetry. Capitalisation, indentation, punctuation, line length, visual parallelism and sectioning will be considered. Example 4.23 (Wordsworth 1983 [1804]: 206) My heart leaps up when I behold A Rainbow in the sky: So was it when my life began; So is it now I am a Man; So be it when I shall grow old, Or let me die! The Child is Father of the Man; And I could wish my days to be Bound each to each by natural piety.

1 2 3 4 5 6 7 8 9

In Example 4.23, the ﬁrst word in every line is capitalised to mark the start of a new line. Because capitalised letters are larger and more noticeable, the repetition of ‘So’ in Lines 3, 4 and 5, is more distinct than the repetition of ‘when’ in the body of the text in Lines 1, 3 and 5. ‘So’ is a pro-form, substituting for the ﬁrst sentence, ‘my heart leaps up’. Hence, the repetition of ‘so’ is semantically as well as visually charged. This is a diagrammatic manifestation drawn from the grammatical metaphor, MORE CONTENT IS MORE FORM.

118 Metaphor and Iconicity

Capitalised words are also used in poetry to express something extraordinary, unique or important. It is a case of DIFFERENCE OF MEANING IS DIFFERENCE OF FORM. In this particular poem, there are several capitalised words, ‘Rainbow’, ‘Man’, ‘Child’, ‘Father’ and ‘Man’ in Lines 2, 4 and 7. Particularly, all the nouns in Line 7 are capitalised. Their emphasis by capitalisation makes this the key line of the poem. Semantic content may also be suggested by indentation; the indented lines, 2 and 6, are thus foregrounded. Indentation, another diagrammatic manifestation of iconicity, distinguishes lines of central and focal importance. ‘A Rainbow in the sky’ is the only concrete image in the poem; ‘let me die’ is an index of the poet’s passion, which is also strengthened by the use of an exclamation mark. The power of blank space should never be taken lightly. When a poem is laid out on the page, it is the blank space which divides the stanzas or sections.16 It visually signals that the stanza (or the section) is a group of words and sentences. The form of the stanza (or the section) suggests a cohesive content, thus representing an abstract diagrammatic correspondence between form and content. This holds true in both English and Japanese poetry. 4.3.2 Orthographical revisions of haiku When a language, such as Japanese, has different writing systems to represent a single phonetic text, how a poem looks is as crucial as how it sounds. Both traditional and modern poets explore various possibilities in the visual presentation of their works. Some make overt attempts; others make covert ones. Haiku texts, written in one vertical line, for example, are the most plain-looking with respect to the graphic shape. And yet as will be seen, they also display rich iconic motivations in visual terms. This section, therefore, looks at different orthographical versions of the same poetic texts in order to speculate as to what motivations are behind the different versions and what visual effects they have. I have chosen as data the revisions of haiku made by Basho Matsuo himself in one of his collections, Oku no Hosomichi (Matsuo (1996 [1694]).17 It is interesting to see how the poet experiments intentionally with possibilities before hitting on the combination of scripts which best represents his conception. I will discuss how the process of changing characters in textual revisions achieves various semantic and structural effects. This analysis concerns particular visual revisions the poet made in changing between kanji and hiragana. As discussed in Section 4.1,

Vision as Meaning 119 Table 4.2 Different types of revisions Types of revision

No. of poems

% out of the total of 48

Orthography Phonology Semantics Syntax

45 31 24 14

94 65 50 29

kanji and hiragana give contrasting visual effects in two major ways: the angularity of the kanji leads to masculine associations such as sharpness, strength and tension, while the cursive shape of the hiragana leads to feminine associations such as gentleness, softness and smoothness; and the integrity of the visual–semantic link is much stronger in the logographic unit of kanji than in the phonographic unit of hiragana. Comparison of several drafts and copies of Oku no Hosomichi shows that Basho revised 48 poems out of 50, and in each of these he tended to revise more than once before arriving at the ﬁnal version. Table 4.2 shows the distribution of different types of revisions, including changes to orthography, phonology, semantics and syntax. The dominance of orthographical revision is obvious as 94 per cent (or 45) of the poems have at least some revisions pertaining to orthography, whereas the corresponding ﬁgures are variably less for the other types of revision. Furthermore, a total of 86 items in 45 poems were orthographically revised. This means that in each poem an average of approximately two items had their character changed. In examining the revisions, three factors have been taken into consideration: the correspondence between the semantic content of a word and the visual association of the character(s) changed; the relationship between the semantic content of the text and the visual association of the character(s) changed; and the positions of the character revisions in the structure of the text. Although these three factors seem to work jointly rather than independently in the visual revisions, I will point out the general tendencies of respective factors in this section.18

Character choice and the meaning of the word Firstly, Basho has a general tendency throughout Oku no Hosomichi to spell the names of creatures in kanji rather than hiragana. Actually 77

120 Metaphor and Iconicity

per cent (23 items out of 30) of the names of living creatures were spelled in kanji in the ﬁnal version. Eight items were revised from hiragana to kanji, and only three items from kanji to hiragana. The compactness and tension of kanji are perhaps more appropriate as the mode of representation of delineated objects like animals, ﬁsh, birds and insects described in Basho’s poems. Secondly, items referring to time and place tended to be written in kanji (80 per cent, 25 out of 29 items) such as names of the seasons, months and places. Similarly, items describing natural phenomena such as the sun, the moon, clouds, wind and rain were written entirely in kanji (14 out of 14) and so were the natural objects such as a waterfall, ﬁelds, rocks, mountains, ocean, beaches, waves and sands (13 out of 14 items). Note that in Oku no Hosomichi, those items written in kanji about time, place and nature are ‘masculine’ objects such as ‘summer’, ‘the sun’, ‘rapid rain’, ‘rough sea’, ‘mountains’ and ‘rocks’. Here, we can point out an iconic correspondence on the image level between the semantic content of the words and the visual association of kanji characters.

Character choice and the meaning of the text Each word is an integral part of the haiku text, and each haiku is an integral part of the entire travel sketch. Thus, it will be fruitful to consider the relationships between the character choice and the meaning of the text, and then between the character choice and the structure of the whole work. Revisions of character types with respect to the verbs offer interesting examples. Although it is a subtle tendency as is seen in 6 items through 6 revisions, Basho tended to revise kanji to hiragana when describing actions such as breaking, collecting, tearing and winning. Interestingly enough, there was no revision in the reverse direction, hiragana to kanji. The revised verbs have semantically more masculine connotations than feminine, which are seemingly contradictory to associations inherent in hiragana. These counter-examples make us wonder if there are some motivations overriding the iconic correspondence between the character choice and the meaning of the words. Examples 4.24 and 4.25 indicate that some of the revised verbs with masculine connotations were used in negation in the text, so that the textual meaning (i.e. negation of masculinity in the action described by the verb) does not contradict the implied connotation of the revision from kanji to hiragana.

Vision as Meaning 121

Example 4.2419 (a) (b) kitsutsuki mo Woodpecker even

iwo wa yabura -zu cottage TOP break-NEG

natsu kodachi summer grove

‘Even woodpeckers don’t tear at the hut in summer trees’ (Matsuo 1996 [1694]: 57) The verb in question, which was revised from 4.24(a) to 4.24(b), is boxed. Note that the verb, yaburu (‘to break, to tear’), is used in a negative form yabura-zu (‘not to be broken’), and the entire text reads ‘even the woodpeckers haven’t pecked a cottage in the summer grove’. It is interpreted as if the woodpeckers did not dare to disturb the peaceful and humble cottage in the serenity of the summer grove. The smoothness and softness associated with hiragana evoke this serenity better than kanji. In Example 4.25, the verb of motion, oru (‘to pluck, to break’), was revised from kanji (4.25(a)) to hiragana (4.25(b)). Example 4.25 (a) (b) Yamanaka ya Yamanaka:

kiku wa taora-nu chrysanthemum TOP pluck -NEG

yu no nioi hot.water of smell

‘In Yamanaka I don’t break mums off in hot water’s scent’ (Matsuo 1996 [1694]: 119). The haiku praises the good smell of hot springs in Yamanaka. Because the scent is so good, it is not necessary to pluck the chrysanthemum for any additional fragrance. A Chinese legend says that the dew on the chrysanthemum is a life-prolonging drink. Alluding to this legend, the poem also compares the effectiveness of the chrysanthemum and the hot water, suggesting that one could live as long by taking this hot bath rather than plucking the chrysanthemum for its dew. The overall semantic tone of the text is gentle, smooth and soft. The verb oru (‘to pluck’) was used in negation, and revised into hiragana. The change of character type is consonant with the semantic content of the text. Notice also that the last word, nioi (‘smell’) was revised from to as

122 Metaphor and Iconicity

boxed in Example 4.25(a) and (b). Both characters are kanji, but have different density and angularity. (4.25(a)) is visually more dense and tense than (4.25(b)). This change also contributes to evoking the soft and gentle association of the poem. The structural meaning of character choice Character choice not only reinforces the meaning of a poem but also affects the visual structure of the poem. This means that the position of kanji or hiragana, the sequential order of the character types and repetition in general, suggest the structural orientations of the poem, which navigate the reading of the text in a certain direction. Possible structural effects derived from the revision of the character are: (1) intensifying; (2) foregrounding; and (3) bracketing. It should be noted that these effects are derived by grammatical metaphors of repetition discussed in Chapter 2 (Sections 2.4.2 and 2.4.3 in particular). (1) Intensifying. Using the same character, or the same character type, repeatedly creates certain effects. Abundance and intensity are notable iconic effects produced by such repetitions as the grammatical metaphors of MORE CONTENT IS MORE FORM and MORE STRENGTH IS MORE FORM (for further discussion on repetitions, see Hiraga 1993). Example 4.26 illustrates the use of repetition of the same character. Example 4.26 (a) (b) ara tooto alas, in awe

aoba wakaba no green leaves fresh leaves of

hi no hikari sun of light

‘Look, so holy: green leaves young leaves in the light of the sun’ (Matsuo 1996 [1694]: 47). The poem describes the beauty of young green leaves in the sunlight. The poet is in awe not only in looking at rich natural beauty, but also in respecting the magniﬁcent temple which houses the Tokugawa Shogun in Nikko. Hi no hikari written mainly in kanji is a visual pun, meaning ‘the light of the sun’ and the place name, Nikko ( ). In order to convey these rich connotations, perhaps the structural revision was preferred. A change occurred in the word wakaba (‘young leaves’) initially written in hiragana (4.26(a)) and revised to kanji (4.26(b)). Connotations of ‘young leaves’ such as a pale green colour, soft texture and tender

Vision as Meaning 123

smell, are all feminine. However, in the revision, the repetition of ‘leaves’ (underlined in Example 4.26(b)) appears to have been preferable to the hiragana spelling despite the obvious iconicity of smoothness and softness evoked by the hiragana, with the semantic content of ‘young leaves’. The structural effect created by the diagrammatic manifestation of the abundance of the ‘leaves’ overrides the correspondence between character choice and the word. The more characters of the same sort ( in this example), the greater the focus on the abundance of the ‘leaves’. Showing ‘more leaves’ iconically leads to a richer appreciation of natural beauty. This change also affects the overall appearance of the poem. Now the ﬁrst ﬁve morae, ara tooto (‘alas, in awe’) in hiragana is put in tension with the latter part of the poem mainly written in kanji. There is a reason for creating this tension. For ara tooto shows the poet’s admiration of the prosperity of the Tokugawa Shogunate, hinted at by the place name, as well as his obvious admiration of the leaves. (2) Foregrounding. Another important structural effect of kanji is foregrounding. As explained earlier, kanji have a distinct form and semantic integrity, which differentiate them visually as the ﬁgure while the remaining hiragana function as the ground. The following haiku describes the grandeur of the Milky Way, which arches over the rough waves of the Sea of Japan towards Sado Island (for a more elaborated analysis of this haiku, see Chapter 1).

Example 4.27 (a) (b) araumi ya rough sea:

Sado ni yoko tou Sado in lie

ama no gawa heaven of river

‘Rough sea: lying toward Sado Island the River of Heaven’ (Matsuo 1996 [1694]: 109). In Example 4.27, words put into semantic tension, i.e. (‘rough sea’) and (‘milky way’), were spelled in kanji. Sado , a place name, is also written in kanji. The main effect of changing the character type in the verb yokotou (‘to lie’) from kanji to hiragana is to make that part of the text a ground for the conspicuous proﬁle of (‘rough sea’), (Sado) and (‘milky way’). Thus, the foregrounded visual elements signal the foregrounding of the meanings expressed.

124 Metaphor and Iconicity

Notice also that the three nouns, araumi (‘rough sea’), Sado and ama no gawa (‘milky way’), are all in two kanji. Also, (‘sea’), (‘to cross water’) and (‘river’) in these three nouns (underlined in Example 4.27) share the same radical signifying water. Both (‘rough sea’) and (‘milky way’) relate to water, because the Milky Way in Japanese is called ‘the river of heaven’. Therefore, the repetitive use of similar visual elements – two-character nouns and the same radical – further intensiﬁes the meaning of the elements visually foregrounded by the use of kanji. (3) Bracketing. Bracketing is a structural effect produced by the use of the same type of linguistic units at the beginning and the end of the text. One may detect here the grammatical metaphor of A FRAME IS SAMENESS OF BEGINNING AND ENDING. For auditory bracketing, a poem may begin with the same sound, syllable, word or syntactic structure (such as question, negation, relative clause, etc.), whereas for visual bracketing, it may begin with the same letter, character or shape as it ends with. Visual bracketing is observed in 68 per cent of haiku in Oku no Hosomichi; 31 (62 per cent) out of 50 poems begin and end with kanji, while 3 poems (6 per cent) begin and end with hiragana. Many of the orthographical revisions by Basho work towards this bracketing effect. It would appear that he consciously used this iconic device. Example 4.28 is an instance where the type of the ﬁnal character is changed from hiragana to kanji to bracket the poem with kanji. Example 4.28 (a) (b) hai-ide yo crawl.out

kaiya ga shita no silkworm.nursery SUB under of

hiki no koe toad of voice

‘Crawl out, toad: your voice under the silkworm shed’ (Matsuo 1996 [1694]: 93). The poem addresses a toad, which is hidden under the silkworm shed. It is hot and humid. So, when the poet hears the voice of the toad, he tells it to crawl out for fresh and cool air. The word koe (‘voice’) was ﬁrst spelled in hiragana, possibly because of the internal structural uniformity with hiki (‘toad’), which was also spelled in hiragana. In the revision, the external structural bracketing perhaps overrode the internal structural uniformity.

Vision as Meaning 125

It is also notable that bracketing with kanji is a characteristic distinctive to the poems in Oku no Hosomichi. Bracketing the poems with the angular and distinctive shape of kanji visually delineates the haiku texts from the body of the travel sketch, and thereby lends them independence. At the same time, the repetition of this structural trait of bracketing in more than 60 per cent of the poems gives a sense of continuity to the haiku texts as a whole.

4.4 Concluding remarks This chapter demonstrates that iconicity is an important motivation in both the construction and interpretation of poetry with special reference to the visual elements in the texts. Some poems display an iconic link between the visual forms and their meanings overtly; but most of the poems show the link more subtly. Visual poetry provides samples of an overt manifestation as it uses iconicity in concrete and exaggerated ways so that the principles of iconicity can be clearly delineated. On the other hand, by looking at certain abstract and hidden patterns in the distribution of visible linguistic elements, the analyses also demonstrate that a discovery of such subliminal iconic links between the form and meaning yields new interpretations of poetic texts. Poets are bound to their own visual resources and the use of visual space on the page. Nonetheless, each poetic text deﬁnes the variety of visual options and the extent to which the poet utilises them. Generally, visual poetry shows more iconic motivations in the use of graphic space as has been seen in the examples of pattern poetry. Concrete poetry and free verse, on the other hand, experiment with the creation of new visible units (especially Examples 4.20 and 4.21) and new conﬁgurations of visual linguistic elements. In conventional poetry, too, poets are conscious about the use of graphic space and the appearance of the poem on the page. There are structural reasons why they use indentation, sectioning and the repetition of certain linguistic patterns to describe certain things, and if we approach the texts with iconicity in mind, we can often explain why. Subtle iconic effects provide insight; once we uncover the link between the visual form and the meaning it expresses, this link often directs or reinforces our interpretation of the text in that way. As has been seen in Basho’s use of kanji logographs in his revisions, visual images and diagrams revealed by the analyses are so powerful that it is often difﬁcult to escape from their spell. This says something about the power of the written language, which has been neglected in modern theories of language and poetry.

126 Metaphor and Iconicity

A comparison of English and Japanese poetic texts in visual terms also generates a greater awareness of the written language, because both languages employ very different writing systems. The writing system of English is dominantly phonographic, while the writing system of Japanese is a mixture of phonographic and logographic units. Firstly, the difference appears in the degree of consciousness about the written language. For the Japanese poets, it is a conscious effort to choose which character type, hiragana, katakana or kanji, to spell out the sounds and the contents they want to represent. Similarly, it is a conscious practice on the part of the reader to read what is represented in the poems not only by the semantic meaning conveyed but also by character choice and layout. Secondly, because kanji characters are distinct in form and integrated in meaning, they are generally perceived as the ﬁgure rather than the ground in the process of reading. Being logographs, many kanji convey their meanings iconically, too.20 Therefore, the choice of a particular kanji to represent what the poets want to convey is also crucial in the making of poetry (for further discussion, see Section 7.2.2). Likewise, experienced readers read implied meanings in the choice of certain kanji characters.21 Consequently, it can be said that Japanese poetry is more conscious about the visual representation than English poetry. It is no wonder that visual iconicity is generally more noticeable and prevalent in Japanese poetry, not only in visual poetry but also in conventional and traditional poetry. The poetic forms, whether they are as immediate as images, as structural as diagrams or as conceptual as metaphors, display iconic elements to varying degrees. This also depends on the nature of the written language and the writing systems that the poetic text employs. Nonetheless, the point is that far from being subordinated to meaning, the visual form of poetic text in both English and Japanese plays a leading role in conveying and creating meaning. Metaphor–icon links manifest themselves at this very point where the form contributes to a reinforcement or creation of meaning.

5 Sound as Meaning

The sound must seem an echo to the sense. Alexander Pope (1709)

5.1 Poetry and auditory representation Sound as well as vision has a rich tradition in poetic analysis. It can constitute the form of a poem, offering shape and material. The quality inherent in sounds gives texture to a poem, whereas the pattern of sounds gives it architecture. Sound may not only be essential in terms of its formal qualities, but also important in terms of its intimate relationship with the meaning of a poem. The relationship between sound and meaning in poetry is an extremely close one. For example, Shapiro and Beum have said that usually a poet ‘creates a form that “says” the same thing as the words themselves say’ (1965: 2). This wedding of sound and sense – how sound contributes to complementing the whole experience of a poem – has been the focus of an enormous amount of endeavour in poetic research. Traditionally it has been studied in the ﬁeld of prosody or metrics. The phonetic analysis in this chapter is based on linguistic poetics, which incorporates traditional metrics with linguistics and semiotics (cf. Jakobson 1960, 1971, 1981, 1985, Jakobson and Waugh 1979, among others). The immediacy of the relationship between sound and sense is reinterpreted by using the concept of ‘iconicity’. After discussing the basic characteristics of the sound systems of modern standard Japanese1 in comparison with those of British English, the chapter will present an analysis of auditory manifestations of the interplay between metaphor and iconicity in some illustrative poetic texts in both languages. It can be speculated that differences in the sound structures may 127

128 Metaphor and Iconicity

affect the kinds of metaphor–icon links manifested in Japanese and English poetic texts. English has more complex sound structures, and thereby has developed a more sophisticated set of phonological devices for poetic creation. Japanese, on the other hand, has relatively simple sound structures and less elaborate phonological devices for poetry. Despite these possible phonological differences, however, I would like to claim the following two points: the ubiquity of metaphor–icon links motivating poetic texts; and the usefulness of the concept of iconicity as a methodological tool for analysis. 5.1.1 The sound systems of English and Japanese Just as poetry exploits various options in the writing system for its visual effects, it also explores possibilities in the sound system to achieve certain auditory effects. Such possibilities include sound qualities inherent in each sound, as well as conﬁgurations or patterns of sounds employed in the text. This section looks at the differences in Japanese and English phonemic and syllabic structures, and relates them to their differing writing systems. More speciﬁc issues concerning the auditory representations of poetry are discussed in the following section with special reference to onomatopoeia, sound-symbolism and general features of rhyme and rhythm. It is not the aim of this chapter, however, to offer an exhaustive comparison of Japanese and English sound systems nor to provide a detailed analysis of their metrics. The purpose is to cover the basic aspects of their differences in order to discuss the auditory manifestations of imagic and diagrammatic iconicity in Japanese and English texts. Sound units in Japanese and English: syllable and mora The essential difference between English and Japanese concerns the sound units, syllable and mora. Although both English and Japanese have phonemes as the smallest sound unit, the way they formulate the units larger than phonemes is different. A mora is a unit of ‘timing’ (Ladefoged 1982: 226), whereas a syllable2 is deﬁned in terms of the inherent sonority of each sound. The sonority of a sound is its loudness relative to that of other sounds with the same length, stress and pitch (Ladefoged 1982: 221). Therefore, the number of peaks of sonority in a word decides the number of the syllables of the word. Generally, vowels have a higher degree of sonority than consonants; and that is where the peaks of sonority in a word often reside.

Sound as Meaning 129

A native speaker of English might think of a word such as Nippon (‘Japan’) having two syllables, [niQ] and [pon]; but a native speaker of Japanese further divides the word into the four units, [ni], [Q ], [po] and [N].3 These four units take about the same length of time to pronounce, and constitute four morae. The most common type of Japanese mora consists of a consonant followed by a vowel (a [CV] mora). Another type is a vowel by itself, as in the word [a-i] (‘love’). This word has two morae. A long vowel is counted as having two morae. So a word such as [o-o-ki-i] (‘big’) has four morae. A consonant cannot occur after a vowel within a mora; but it too can form a mora by itself. A glottal stop [Q ] and a moraic nasal [N] are of this kind. Therefore, [V] in a [V-V] sequence, [Q ] and [N] are all pronounced with the same length of time as a [CV] mora.4 The unit of mora is of particular importance in the discussion of Japanese poetic texts in two ways: its relation to the Japanese writing system; and its function as a rhythmic unit. Firstly, each mora in Japanese is represented by one letter in kana syllabaries. Unlike an alphabetic writing system in which syllables are systematically split into their components, in the Japanese syllabaries there is one letter for each mora in each of the hiragana and katakana writing systems. Hence, a one-mora word like [ki] (‘tree’) can be written in one letter such as (in hiragana) or (in katakana); and a two-mora word like [i-ki] (‘breath’) in two letters such as (in hiragana) or (in katakana). It is widely observed that people who have not been educated in an alphabetic writing system ﬁnd it more difﬁcult to consider syllables as being made up of segments such as consonants and vowels (Ladefoged 1982: 220). This is the case with Japanese speakers, who do not consciously decompose the morae into smaller segments when they speak or hear the sounds in their natural utterances. Indeed, the mora, being a unit of both speech and writing, seems to be a dominant unit in the mental organisation and production of utterances in Japanese. Secondly, the mora is important in the rhythm of poetic texts. Japanese poetic texts are mora-timed, whereas English texts are stress-andsyllable-timed. The mora-timed texts have a rhythm in which each mora tends to have the same duration. The most prominent poetic forms in Japanese literature, haiku and tanka, have 17 and 31 morae respectively, with an alternation of the 5-mora line and the 7-mora line. This rhythm pattern is called goshichi-cho (‘5–7 melody’) or shichigo-cho (‘7–5 melody’), and is dominant not only in poetry but also in copyscripts, slogans and popular song lyrics.

130 Metaphor and Iconicity

By contrast, in English, the number and the conﬁguration of stressed and unstressed syllables are important factors in metrics. In other words, it is a combination of the unit of the syllable and the suprasegmental phenomenon of stress that determines the rhythm of English poetry in which stressed syllables tend to recur at regular intervals of time. In short, both English and Japanese have a speciﬁc way of measuring the poetic rhythm unit. English uses most predominantly the stressed syllable, while Japanese uses the mora. The crucial difference between the two languages seems to be in the perception of individual sounds or sound units in relation to the writing systems. Using an alphabetic writing system, English speakers ﬁnd it comparatively easy to identify syllables and their sound make-up, whereas Japanese speakers, largely inﬂuenced by their kana syllabaries, are more tuned in to hearing the mora sequence rather than the sequence of each sound. Here again, the evidence from Japanese indicates that the written language can play a role in guiding the auditory perception of literate speakers, and this is particularly dominant in reading and composing poetic texts in Japanese (cf. Shibatani 1990: 158). The phonemic structures of Japanese and English Having said that the unit of the mora and syllable is perhaps more crucial than the unit of the phoneme when Japanese texts are compared to those of English, however, it is worth discussing brieﬂy the phonemic structures of Japanese and English. As an inherent quality of sounds, phonemes relate closely to the sound materials of poetic texts, particularly onomatopoeia and sound-symbolism. They also count in terms of the quantity of the auditory materials available in both languages. Japanese and English differ in the number of phonemes each language has. English has a greater variety of vowels and consonants. Japanese has only ﬁve vowels, [i], [u], [e], [o], [a], while English has more than a dozen. Compared to 18 consonants in Japanese, English has 22.5 English and Japanese differ not only in the number of phonemes, but also in the make-up of syllables or morae. As discussed above, Japanese morae have rather simple constructions, compared to those of English syllables in which English allows consonant clusters, diphthongs, long vowels, and so on. In Japanese, most words end with a vowel, whereas English words can end either with a vowel, a consonant or a combination of consonants. English has three times as many vowels as Japanese. This is perhaps a major differing factor in the rhyming potential of each language. For example, Japanese traditional poetry does not seek to effect assonance

Sound as Meaning 131

or rhymed feet in the way that European poetry does. Alliteration can occur in Japanese poetry; but, again inﬂuenced by moraic perception, it is not as strongly felt as in European poetry. It can be concluded that the major phonological differences between Japanese and English are drawn from two sources: the sound unit of mora; and the inventory of phonemes. With regard to both sources, Japanese has a far simpler sound structure than English. My tentative prediction from this conclusion is that English poetry perhaps exploits phonological resources and effects more readily and widely than Japanese poetry. 5.1.2 Auditory representation in poetry It has been claimed that poetry exploits the resources of sound more subtly and more thoroughly than any other kind of discourse to produce effects which often reinforce the meaning (e.g. Shapiro and Beum 1965, Leech 1969, among others). The devices utilised for such auditory effects include onomatopoeia, sound-symbolism, rhyme and rhythm (metre). The former two devices concern the inherent qualities of sounds, whereas the latter two concern the conﬁguration of sound units. This section attempts to offer a new outlook on these devices from the perspective of auditory iconicity. Onomatopoeia and sound-symbolism In a rigorous analysis, onomatopoeia and sound-symbolism are different in nature: onomatopoeia occurs when speech sounds imitate other physical sounds; and sound-symbolism when sounds are associated with other sensual or conceptual dimensions. Onomatopoeia occurs most typically in words, whereas sound-symbolism occurs in both words and in particular sounds, or more technically in distinctive features. Onomatopoeia can be understood in two major ways. Its narrowest and most literal deﬁnition is the purely mimetic use of speech sounds to describe other non-linguistic physical sounds. This deﬁnition of onomatopoeia includes, for example, the ‘cries’ of animals and birds, the natural sounds of wind and raindrops, the mechanical sounds of objects such as clocks, cars, trains, and so on. A wider interpretation of onomatopoeia concerns the phonological patterns that represent not just the sound of what they describe but the activity as a whole. For example, when Tennyson says, ‘the murmuring of innumerable bees’, the repetitive pattern of bilabials and nasals, [m], [n] and [b], connotes the humming sound of the bees and an activity of countless bees (cf. Wellek

132 Metaphor and Iconicity

and Warren 1963[1949]: 162) together with the onomatopoeic word, ‘murmuring’. This second interpretation of onomatopoeia is perhaps an area in which a clear division from sound-symbolism is difﬁcult; however, this tends to be the area that poets are more interested in exploring than the narrow sense of onomatopoeia. Sound-symbolism, ‘the tendency of particular sounds, usually irrespective of the language in which they occur, to correlate with aspects of the meaning of words containing these sounds’ (Kakehi et al. 1996: xxii) is far more exploited in poetry than onomatopoeia, possibly because it has more subtle effects and because it can appear in individual sounds as well as in words rather than just in particular words alone. Shapiro and Beum (1965: 10–12) categorise the possible soundsymbolism of consonants utilised by English poets as below. This categorisation can also apply to Japanese poetry:6 1. Resonance. In the present book we use this term stipulatively to mean prolongation and fullness of sounds. The sounds n, m, N , z and Z (none, maim, ring, nose, rouge) usually produce lingering droning, vibrant effects. The nasals (n, m, N ) are sometimes called hums. 2. Harshness. Includes throaty sounds, usually called velars, made by putting the tongue back toward the soft palate: k and hard g, as in kick and jug; or hard k, as in cough. Synonyms are dissonance and cacophony. 3. Plosiveness. Sounds articulated by a sudden release, and then an interruption, of breath: b, p, t, d, g and k are often plosive. 4. Breathiness. There is a wide range of sounds – it includes the groups known to linguists as aspirates, sibilants and fricatives – that are especially suitable in contexts that seek to create images of such qualities as breathlessness or hissing or whispering. The aspirate h perhaps should be considered separately. It may be employed to create a suggestion of breathlessness, as in fatigue or wonder or exaltation. 5. Liquidity. This is the effect of certain non-frictional and vowel-like consonants, chieﬂy l and r. W is another mild consonantal sound. These melodic consonants can help a poet build a resonant musical line. They are frequently conjoined with resonants. . . . The voiced sounds l, r, y, w, m and n may be grouped together and called sonorants. Leech (1969: 98–9) also lists classes of English consonants on a scale of increasing hardness:

Sound as Meaning 133

1. 2. 3. 4.

Liquids and nasals: /l/, /r/, /n/, /N/ (as in ‘thing’). Fricatives and aspirates: /v/, /Q/ (as in ‘there’), /f/, /s/, etc. Affricates: /tS/ (as in ‘church’), /dZ/ (as in ‘judge’). Plosives: /b/, /d/, /g/, /p/, /t/, /k/.

The presence of voice is another factor which affects soundsymbolism. Voiced consonants (/v/, /Q/, /z/, etc.) have a more relaxed articulation than their voiceless counterparts (/f/, /q/, /s/, etc.). They tend to suggest softness in English, whereas in Japanese voiced consonants tend to indicate one or more of the following effects when used in describing the manner of actions or states (Kakehi et al. 1996: xxiv): 1. 2. 3. 4. 5.

larger size greater quantity or volume more rigorous actions more intense degree of action or state negative meanings or associations, caused by the association of excessiveness in terms of 1 to 4 above.

Vowels are voiced sounds and have a great preponderance over consonants. In particular, long and open vowels and long diphthongs often produce a kind of softness somewhat different in quality from the softness that may be created by resonant or breathy or liquid consonants. [a] and [o] tend to suggest largeness and slowness, while [i] has the opposite tendency. These onomatopoeic and sound-symbolic factors play an important role in the imagic mapping of the auditory form onto the meaning. It seems that each sound-symbolic factor can be mapped onto different spaces in the meaning content of the text. This point will be elaborated further in the analysis of each poetic text in this chapter. Rhyme and rhythm It has often been claimed that there is a rich array of techniques in poetry that enhance auditory iconicity in terms of images and diagrams. The distinction between imagic iconicity and diagrammatic iconicity corresponds neatly with the inherent element and the relational element of sound in poetry (cf. Wellek and Warren 1963 [1949]: 159). The former is the peculiar individuality of the sound, or an independent quality inherent in the sound itself, whereas the latter is determined in relation to other sounds, for example the pitch, the duration of the sounds, the stress, the rhyme, the repetition, and so on, which are the basis of rhythm and metre.

134 Metaphor and Iconicity

In contrast to imagic iconicity, auditory diagrammatic iconicity is prevalent in many poems, in which similarity relations in the phonological form are used as cues to the meaning through rhyme, semi-rhyme, alliteration, assonance, metre and all the other kinds of repetition of sounds, syllables, words, and so on. When words are brought together by rhyme, linked up or contrasted, it is often the case that readers are navigated by the sound make-up of the rhyme to read in the text something similar, something linked up or something contrasted into the text.

Summary In summary, the differences in the sound structures covered in this section affect the possible phonological representations manifested in Japanese and English poetic texts. Obviously English, having a greater number of sounds and more ways of combining these sounds, has developed more sophisticated and elaborate phonological devices for poetry than Japanese, which has a far lesser number of sounds and their combinations. Hence, poetic forms are phonologically radically different in the two languages. Having said that English and Japanese have these differences in phonological resources, however, I would like to show in the following sections that poetic texts in both languages can be analysed by the same notional tools of metaphor and iconicity, and that both texts display the interplay between metaphor and iconicity quite fully and in a variety of sophisticated ways, within the constraints of their own languages.

5.2 Phonosymbolism This section presents two illustrations of auditory manifestations in which imagic iconicity predominates over diagrammatic iconicity. The ﬁrst example, Edgar Allan Poe’s poem ‘The Bells’, demonstrates that the auditory form of a poem exhibits its meaning by the use of onomatopoetic words, sound-symbolism in several kinds of repetitive sound forms and syntactic parallelism. The second example is a haiku by Basho. It also utilises sound-symbolic effects. The analysis will look at the revising process of the poem to demonstrate how the poet came to achieve the auditory iconicity. In both examples, the abundant use of onomatopoeia and sound-symbolism illustrates how imagic iconicity is successful in these texts. Diagrammatic iconicity is more indirect and

Sound as Meaning 135

abstract than imagic iconicity, and it is less prominent in the texts. Nevertheless, the sample analyses aim to demonstrate that both types of iconicity are intertwined with the metaphorical structuring of the text to evoke the intended meaning effectively. 5.2.1 The dance of sounds in Edgar Allan Poe’s ‘The Bells’ ‘The Bells’ by Edgar Allan Poe (1919: 63–6) is known for its clever use of onomatopoeia and sound-symbolism.7 The following analysis will brieﬂy demonstrate how and to what degree auditory iconicity is prevalent and crucial in the sound make-up of this poem.8 Example 5.1 THE BELLS I HEAR the sledges with the bells – Silver bells! What a world of merriment their melody foretells! How they tinkle, tinkle, tinkle, In the icy air of night! While the stars that oversprinkle All the Heavens, seem to twinkle With a crystalline delight; Keeping time, time, time, In a sort of Runic rhyme, To the tintinnabulation that so musically wells From the bells, bells, bells, bells, Bells, bells, bells – From the jingling and tinkling of the bells.

1 2 3 4 5 6 7 8 9 10 11 12 13 14

II Hear the mellow wedding bells – Golden bells! What a world of happiness their harmony foretells Through the balmy air of night How they ring out their delight! – From the molten-golden notes, And all in tune, What a liquid ditty ﬂoats To the turtle-dove that listens, while she gloats On the moon!

1 2 3 4 5 6 7 8 9 10

136 Metaphor and Iconicity

Oh, from out the sounding cells, What a gush of euphony voluminously wells! How it swells! How it dwells On the future! – how it tells Of the rapture that impels To the swinging and the ringing Of the bells, bells, bells – Of the bells, bells, bells, bells, Bells, bells, bells – To the rhyming and the chiming of the bells! III Hear the loud alarum bells – Brazen bells! What a tale of terror, now, their turbulency tells! In the startled ear of Night How they scream out their affright! Too much horriﬁed to speak, They can only shriek, shriek, Out of tune, In a clamorous appealing to the mercy of the ﬁre, In a mad expostulation with the deaf and frantic ﬁre Leaping higher, higher, higher, With a desperate desire, And a resolute endeavour Now – now to sit, or never, By the side of the pale-faced moon, Oh, the bells, bells, bells! What a tale their terror tells Of despair! How they clang, and clash, and roar! What a horror they outpour On the bosom of the palpitating air! Yet the ear, it fully knows, By the twanging, And the clanging, How the danger ebbs and ﬂows; Yes, the ear distinctly tells, In the jangling, And the wrangling,

11 12 13 14 15 16 17 18 19 20 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Sound as Meaning 137

How the danger sinks and swells, By the sinking or the swelling in the anger of the bells – Of the bells – Of the bells, bells, bells, bells, Bells, bells, bells – In the clamor and the clangor of the bells! IV Hear the tolling of the bells – Iron bells! What a world of solemn thought their monody compels! In the silence of the night, How we shiver with affright At the melancholy menace of their tone! For every sound that ﬂoats From the rust within their throats Is a groan. And the people – ah, the people – They that dwell up in the steeple, All alone, And who, tolling, tolling, tolling, In that mufﬂed monotone, Feel a glory in so rolling On the human heart a stone – They are neither man nor woman – They are neither brute nor human – They are Ghouls: – And their king it is who tolls: – And he rolls, rolls, rolls, Rolls A Pæan from the bells! And his merry bosom swells With the Pæan of the bells! And he dances and he yells; Keeping time, time, time, In a sort of Runic rhyme, To the Pæan of the bells – Of the bells: – Keeping time, time, time, In a sort of Runic rhyme, To the throbbing of the bells –

29 30 31 32 33 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

138 Metaphor and Iconicity

Of the bells, bells, bells – To the sobbing of the bells: – Keeping time, time, time, As he knells, knells, knells, In a happy Runic rhyme, To the rolling of the bells – Of the bells, bells, bells: – To the tolling of the bells – Of the bells, bells, bells, bells, Bells, bells, bells – To the moaning and the groaning of the bells.

34 35 36 37 38 39 40 41 42 43 44

The semantics of the poem Each stanza, with a distinct theme, describes different types of bells ringing in the night. The ﬁrst stanza, the shortest one, is a description of silver bells which decorate the sledges. The scene is Christmas – merry and delightful. The night air is icy and crystal clear. The sound of the bells is musical and light, mysterious and ancient. The second stanza represents the golden bells of a wedding. The voluminous and harmonious sound of the bells ﬂoats in the warm and gentle air of the night. It is happiness, delight and rapture that the mellow golden bells bring to the beloved mate. The third stanza, on the other hand, conveys quite a different theme. It deals with the brazen alarm bells in the horrifying night. The screaming and shrieking sound of the bells startles the ear, warning about the ﬁre, the terror and the danger. It is the turbulence, the cacophony that is distinct in this stanza. The last stanza refers to the mourning bells. They are the iron bells of solemnity and melancholy. Mysterious monody is tolled by the evil demon in the quiet night. The demon dances and yells to the joy and happiness, while the bells moan and groan their pensive monotone. These themes are stated in the beginning lines of each stanza in a parallel way, and then elaborated further by descriptions of the sounds of the bells. Thematic progression of the stanzas is paraphrasable by the colour of the bells and the mood of their melody as silver, golden, brazen and iron, and merry, happy, terriﬁed and solemn. The poem can also be interpreted as a global metaphor describing the course of human life, such as birth, marriage, life changes and death, by way of the sound of the bells. A schematic representation of metaphor–icon links in ‘The Bells’ is shown in Figure 5.1, to facilitate the discussion on imagic and diagrammatic iconicity which follows.

Sound as Meaning 139

Iconic Mapping

Sound Shape

Metaphorical Mapping

Source sounds of the bells

Target course of life

onomatopoeia IMAGIC MAPPING sound-symbolism parallelism DIAGRAMMATIC MAPPING

Blend FORM

MEANING

Grammatical Metaphors Figure 5.1 Metaphor–icon links in ‘The Bells’

Imagic iconicity As discussed in the previous section, imagic iconicity manifests itself most notably in onomatopoeia and sound-symbolism. ‘The Bells’ is remarkable in this respect with a maximum use of them. It seems as if the poet did an experiment with speech sounds mimetic to the meanings he wanted to infuse in this poem. I will list the onomatopoeic and sound-symbolic elements which are consonant with the themes of the poem. Each stanza contains onomatopoeic and sound-symbolic words9 which convey distinctively different effects and moods of the bells. Table 5.1 is a list of onomatopoeic and sound-symbolic words in each stanza. It should be noted that almost all the onomatopoeic and soundsymbolic words appear in the parallel lines in each stanza, which describe the sound of the bells.10

140 Metaphor and Iconicity Table 5.1 Onomatopoeic words Stanza I

II

III

IV

tinkle (4) [over]sprinkle (6)

ring (5) swinging (17) ringing (17) rhyming (21) chiming (21)

scream (5) shriek(7) clang (19) clash (19) roar (19) twanging (23) clanging (24) jangling (27) wrangling (28) sinking (30) swelling (30) clamor (43) clangor (34)

groan (9) throbbing (33) sobbing (35) rolling (39) tolling (41) moaning (44) groaning (44)

twinkle (7) tintinnabulation (11) jingling (14) tinkling (14)

Table 5.2 Sound-symbolic elements Stanza I

II

III

IV

[i] [t] [. . . inklN] [tin] [. . . ink(g)liN]

[i] [ai] [l] [iNiN] [. . . aimiN]

[æ] [k][Ú][N] [. . . æNiN] [. . . æNliN] consonant clusters

[O] [b] [. . . ObiN] [. . . O:liN] [. . . O:niN]

The sound-symbolic elements characteristic to each stanza are illustrated in Table 5.2. These sounds, either individually or collectively, suggest certain iconic impressions or interpretations to the reader.11 Each stanza differs in suggesting sound-symbolic effects. The ﬁrst stanza is characterised most prominently by the frequent use of [i] and [t]. These sounds share a [-Compact] feature,12 i.e. diffuseness, in common. Diffuse sounds produce symbolic effects of smallness in size and length. The [-Voiced] feature of [t] suggests the crisp and metallic sound. Another notable characteristic is the conﬁguration of a short vowel [i] interrupted by stop consonants such as [inkl], [tin] and [. . . ink(g)iN]. This conﬁguration gives an impression of rapid movement.

Sound as Meaning 141

The second stanza has somewhat similar acoustic qualities – the use of [i]’s, [t]’s and [n]’s. However, the most notable characteristics are the use of two prominent vowel qualities, a short front vowel [i] and a diphthong [ai]; and a frequent occurrence of a liquid sound [l]. In particular, the latter factor contributes to creating a resonant, melodious and smooth image. It should also be noted that this stanza contains more [+Voiced] consonants than [-Voiced], marked by such consonants as [b] and [d], giving an impression of being ‘soft’ and ‘mellow’ rather than ‘sharp’ and ‘hard’. The third stanza is quite different from the former two. The most remarkable feature is [+Compact], namely, a front low vowel [æ] and velar consonants [k], [S] and [N]. As opposed to diffuse sounds, compact sounds give an impression of being large. Another prominent characteristic of the third stanza is the overwhelming use of consonant clusters such as [br], [ﬂ], [fr], [kl], [ks], [kt], [lz], [skr], [sl], [sp], [Sr], [st], [sw], [tl], [tw], etc., which suggests dissonance and cacophony. The repetition of nasals, particularly [n] and [N], also gives a sonorous image. The last stanza, the longest and most repetitive one, is characterised by [+Grave] sounds, such as a low back vowel [O] and a voiced bilabial consonant [b]. This feature suggests a deep, slow and sonorous effect. The back vowels [O] and [o] are elongatable, so that they amplify the length of this stanza.13 The repetitive use of sonorants – nasals, glides and liquids – give an excessively sonorous impression. Also the repetitive use of the same words exaggerates the monotone image. These effects produced by the sound-symbolic elements of sounds contribute to implementing the following moods in each stanza: 1. The continuous sparkling of the light, short, small and rapid sounds of the ﬁrst stanza create happy, merry and joyful moods and a crisp, clear, bright and cold atmosphere; 2. The melodious ﬂow of the soft and smooth sounds of the second stanza create happy, loving and fulﬁlling moods in a joyous, rich and warm setting; 3. The tuneless noise of the big and loud sounds of the third stanza create frightening, disturbing, angry and chaotic moods with a violent and noisy scene; 4. The repetitive monotone of the deep, slow and excessively sonorous sounds of the fourth stanza create sad, lonely, solemn and melancholy moods in an other-worldly atmosphere. Consequently, it can be assumed that the overall auditory mimicry is the repetition of the various bell sounds coloured by different vowels,

142 Metaphor and Iconicity

consonants and their combinations. The total picture of the poem is represented by an overwhelming repetition of the word ‘bells’ and its variants ending with a sound sequence [elz] as well as by the recurring use of gerunds which end with a nasal sound [N]. The [l] sound suggests a ﬂow of continuous movement, and [z] and [N] produce lingering and vibrant effects. [b] and [l], contained in the keyword [belz], are the consonants of most frequent occurrence. It is as if these sounds constituted a uniﬁed undercurrent of the poem – the continuous strokes of the bells – which gets further characterisation by the different qualities of sounds in each stanza. The thematic progression of the stanzas, such as birth, marriage, life changes and death, is consonant with the auditory effects and moods expressed in each stanza. The cacophony plays its fullest effect in the third stanza where we have progressed from innocent childhood and happy youth to an adventurous middle age, whereas the monody characterises the last stanza where life is a slow long routine of an old age. The gaiety and the lightheartedness of the ﬁrst two stanzas are suggested by short, high and rapid movements of the melody. The theme of the terrifying alarm bells of the third stanza is reinforced by the cacophony of loud and disturbing sounds and the solemnity of the last stanza is sustained by the monotonous repetitions of deep and slow sounds. Thus, it has been demonstrated that the effects and the moods manifested by the characteristic sounds of each stanza are iconic to the semantics of the stanza. In other words, the qualities of these sounds display mimetically the psychological reactions brought about by the lexical and textual meanings as well as the stanzaic themes of the poem. In this sense, onomatopoeic words and sound-symbolic features function as images.

Diagrammatic iconicity In the auditory aspect of a poetic text, diagrammatic iconicity occurs in the relation between the sound structure and the semantic structure of a whole or a part of a poem. What follows will look at how the sound structure of ‘The Bells’ manifests diagrammatic iconicity with its themes and meanings. There are several notable sound patterns which need explanation: 1. Each stanza has a similar syntactic framing at the beginning and the end, which contains nouns and verbs, either alliterated or rhymed. The framing format is shown in Figure 5.2.

Sound as Meaning 143 Hear … bells – ADJECTIVE bells! What a NOUN1 of NOUN2 their NOUN3 VERB1 … … …

Framing

PREPOSITION the bells, bells, bells, bells Bells, bells, bells – PREPOSITION the NOUN4 and the NOUN5 of the bells Figure 5.2 Syntactic framing format

Table 5.3 Alliteration and rhyme in the framing lines Stanza

NOUN1 NOUN2 NOUN3 VERB1 NOUN4 NOUN5

I

II

III

IV

world merriment melody foretells jingling tinkling

world happiness harmony foretells rhyming chiming

tale terror turbulency tells clamor clangor

world thought monody compels moaning groaning

In this format, NOUN2 and NOUN3 are alliterated in Stanzas I, II and III; VERB1 is rhymed as [elz] with the ends of the previous two lines, [belz]; and NOUN4 and NOUN5 are rhymed in every stanza and alliterated in Stanza III. Table 5.3 illustrates the alliteration and rhymes in the framing lines. The sameness of the syntactic patterning with similar phonological characteristics signals the sameness of semantic importance. As discussed earlier, these framing lines play a crucial role in describing the theme of each stanza. The distinct alliteration and rhyming in the parallel positions emphasise iconically the importance of the meaning expressed by these accentuated words.

144 Metaphor and Iconicity Table 5.4 Repeated lexical items Stanza

bells tinkle time higher tolling rolls knells

I

II

III

IV

10 3 3

13

15

24 9

3 5 4 3

2. The diagrammatic analogy holds not only on the fairly abstract realisation as above, but also on more accessible levels as the rest of the discussion will disclose. For example, there are lexical items which are repeated more than three times successively in each stanza, as shown in Table 5.4. Their distinct repetition conveys a distinct meaning: the more repetitions of the same words, the more monotonous the effect. It is coherent that the word ‘bells’ is repeated most often as the keyword of this poem. It is also obvious that the fourth stanza contains the largest number of repetitions, and this represents the monody of the bells described in the fourth stanza. 3. There is another obvious repetition of a form throughout this poem, i.e. the morpheme ‘-ing’. Although the words with an ‘-ing’ ending in the list vary in terms of their grammatical category, such as verbs, adjectives or nouns, they all convey an implication of something in progress, a sense of continuity, originally associated with the progressive form of a verb. As seen before, the sound [N] of this morpheme is one of the sonorous and resonant phonemes, which can be interpreted as an embodiment of the resonant bell sounds themselves. Hence, the repetitive use of the ‘-ing’ morpheme suggests iconically the continuous ﬂow of the sounds of the bells. The bell sounds are heard constantly from the beginning to the end of the poem. As the length of the stanzas increases from the ﬁrst to the fourth stanza, there are more occurrences of the morpheme, as if there were more and longer repetitions of bell sounds. As the poem progresses, the colour and the mood of the bells change, which are marked most characteristically by the onomatopoeic words ending with an ‘-ing’, and explained as imagic iconicity. Here, however, diagrammatic iconicity is at work by the navigation of grammatical metaphors: SIMILARITY IN

Sound as Meaning 145

(i.e. a sense of continuity in the repetitious bell sounds) IS (i.e. the ‘-ing’ morpheme); DIFFERENCE IN MEANING (i.e. variation of bell sounds, distinctively different in each stanza) IS DIFFERENCE IN FORM (i.e. different words with which this morpheme is bound). 4. Alliteration is used iconically to bind important words together to secure the intense unity of meaning. Apart from the repeated lexical items mentioned above, there is only one alliteration in Stanzas I and II, nine in Stanza III and six in Stanza IV. For the ﬁrst two stanzas, alliterated words are placed on the framing lines as explained earlier. The last stanza has alliterated passages such as ‘melancholy menace’(6), ‘mufﬂed monotone’(14), ‘human heart’(16), ‘Runic rhyme’ (28, 32, 38), which evoke the theme of mysterious monody. In all the three stanzas I, II and IV, only three sonorants, [m], [h] and [r], are used in alliteration, giving sonorous impressions. In contrast, Stanza III contains a far greater number of alliterations, nine of them, such as ‘brazen bells’(2), ‘tale of terror . . . turbulency tells’(3), ‘they . . . their’(5), ‘frantic fire’(10), ‘desperate desire’(12), ‘now – now . . . never’(14), ‘tale . . . terror tells’(17), ‘clang . . . clash’(19), ‘sink . . . swell’(29, 30) and ‘clamor . . . clangor’(34). All use obstruents such as [b], [t], [Q], [f], [d], [k] and [s] except for one sonorant [n]. The frequent use of alliteration with obstruents is iconic to the meaning of the stanza – attracting attention by the noisy cacophony. 5. The end rhyme scheme of each stanza has some correspondence with the meaning, too. The end rhyme functions to combine words of different semantic contents and different grammatical categories together. The auditory similarity leads one to read, for example, connections between ‘the bells’ and the verbs such as ‘foretell’, ‘well’, ‘swell’, ‘dwell’, ‘tell’, ‘impel’, ‘compel’, ‘knell’, and so on. In this poem, the most prominent set of such rhymed endings is ‘night’ and ‘delight’ in the ﬁrst two stanzas and ‘night’ and ‘affright’ in the latter two stanzas. This represents, in part, a sharp contrast in theme between the ﬁrst half and the latter half of the poem. MEANING

SIMILARITY IN FORM

Concerning diagrammatic iconicity, there is one more interesting fact about the end rhyme scheme of this poem: namely that the syntactic categories and the semantic values of a set of rhymed words which appear exclusively in one stanza display the stanzaic theme. In Stanza I, ‘tinkle’(4), ‘oversprinkle’(6) and ‘twinkle’(7) rhyme with [inkl]. They are all verbs, signifying an active and energetic movement, which are consonant with the theme of this stanza. Stanza II has no such set of

146 Metaphor and Iconicity

rhymed words. In Stanza III, there are six sets of rhymed ends unique to this stanza: ‘speak’(6) and ‘shriek’(7); ‘fire’(9, 10), ‘higher’(11) and ‘desire’(12); ‘endeavour’(13) and ‘never’(14); ‘despair’(18) and ‘air’(21); ‘roar’(19) and ‘outpour’(20); and ‘knows’(22) and ‘ﬂows’(25). The grammatical categories vary, including verbs, nouns, an adjective and an adverb, as if syntactically there were no harmonious agreement in the rhymed words unique to this stanza – chaos – which is one of the thematic characterisations of the stanza. Also the meaning of such rhymed words as ‘shriek’(7), ‘ﬁre’(9, 10), ‘despair’(18) and ‘roar’(19) are extraordinary in the interpretation of this stanza. The last stanza contains three sets of unique end rhyme: ‘tone’(6), ‘groan’(9), ‘alone’(12), ‘monotone’(14) and ‘stone’(16); ‘people’(10) and ‘steeple’(11); and ‘woman’(17) and ‘human’(18). All are nouns except one, which is an adverb. This gives a static implication, corresponding to the monotone of the bells represented by the stanza. It is clear that diagrammatic iconicity is displayed more subtly and subliminally. It can be said that the diagrammatic correspondence supplements a more immediate imagic link between sound and sense. It is sometimes difﬁcult to draw a clear dividing line between images and diagrams, because both work in a complementary manner to produce uniﬁed, dense and rich layers of poetic experience. 5.2.2 Charging the music: phonological revisions of haiku The following haiku is one of the most famous and the most beloved in Japanese literature. It will be claimed that the greater degree of iconicity is achieved through the revisions and, therefore, that an analysis of the revising process shows up the importance of iconicity. Semantics of the text Example 5.2 Shizukasa ya Silence:

iwa ni shimiiru rock in seep.pierce14

semi no koe cicada GEN voice

‘Quietness: seeping into the rocks, the cicada’s voice’ (Matsuo 1996 [1694]: 95). This haiku essentially describes the profound silence and stillness that prevail in nature, and which can also be attributed to the poet himself. Basho intensiﬁes the still silence that is his principal theme by introducing a touch of vivid action, the voice of the cicada. The voice may

Sound as Meaning 147

ﬁrst seem to disturb the moment, but then serves to enhance a still deepened mood of quietness in the poet’s mind. Although the cicada’s voice pierces the rocks, it is an overwhelming silence that is most resonant with the tranquillity of the rocks and the eternal loneliness of the poet. The metaphorical word, shimiiru (‘seep.pierce’), represents this whole process of becoming still. This local metaphor of shimiiru entails the cognitive metaphors, VOICE IS LIQUID and SILENCE IS ROCK. The cicada’s voice and the rocks are put in tension by way of the metaphorical use of shimiiru. The voice is conceptualised as a liquid,15 while the rocks appear to absorb these voices and so to recapture the silence. Indeed, the two natural entities of the poem, rocks and the cicada, are metaphors for silence and sound. This is not the only metaphorical interpretation, however. The cicada, with its short busy life, can also be a metaphor for human beings. There is a cultural cognitive model in Japan in which human life is regarded as being short and transient like the cicada’s life. In contrast, rocks represent the solid and eternal backdrop to human existence. The voice of the cicada seeping into the rocks thus evokes the concept that eternity embraces all transient lives, just as silence absorbs their voices. Indeed, as will be made clear in the more detailed analysis below, the whole text can be interpreted as a global metaphor for the fusion of sound and silence. Using the kireji ‘ya’ as a ‘cutting letter’, the ﬁrst part of the poem, shizukasa (‘silence’), is separated from and contrasted with the rest, iwa ni shimiiru semi no koe (‘the cicada’s voice seeping into the rocks’). In other words, silence is metaphorically equated with the cicada’s voice seeping into the rocks. Basho’s clever use of soundsymbolism or imagic iconicity reinforces the metaphor of eternal stillness, the fusion of sound and silence. In turn, the very metaphor of fusion navigates the diagrammatic interpretation of the text. By considering the poet’s revisions, the following analysis shows how he succeeded in charging his words to create additional poetic power through metaphor–icon links. It also demonstrates that the grasp of such eternal stillness is best represented in the ﬁnal version by its strongest iconic link between the phonological and the semantic aspects of the text.

Revised versions Before Basho arrived at the ﬁnal revision quoted above, it has been shown that he revised it three times as follows.

148 Metaphor and Iconicity

Example 5.3 (Matsuo 1972: 169) (1) ‘Sora’s Draft’ (1689) yamadera ya iwa ni shimitsuku [yamadera ya iwa ni Simitsuku mountain temple: rock to seep.stick

semi no koe semi no koe] cicada of voice

(2) ‘Kogarashi’ (1695) sabishisa no iwa ni shimikomu [sabiSisa no iwa ni Simikomu loneliness SUB seep.include

semi no koe semi no koe]

(3) ‘Hatsusemi’ (1696) sabishisa ya iwa ni shimikomu [sabiSisa ya iwa ni Simikomu loneliness:

semi no koe semi no koe]

(4) ‘Sora’s and Soryu’s Copies’ (1694) shizukasa ya iwa ni shimiiru [Sizukasa ya iwa ni Simiiru stillness: seep.pierce

semi no koe semi no koe]

Although Basho’s revision involved all linguistic aspects, including lexical, syntactic, phonological and orthographical, this section will just cover the lexical and phonological revisions.16 Lexical revisions Basho wrote as follows before introducing the poem: The mountain was made of rocks piled upon boulders, the pines and cypresses were aged, and with the soil and stones old and smooth with moss and the doors of the lesser halls upon the rocks all closed, we heard not a sound. As we went around the cliff, crawled up the rocks, and paid respects to the Buddhist sanctum, the splendid scenery was so hushed and silent that we could only feel our hearts grow clear. Shizukasa ya

iwa ni shimiiru

semi no koe

‘Quietness: seeping into the rocks, the cicada’s voice’ (Matsuo 1996 [1694]: 93–5). As Basho himself so eloquently describes, the focus of his poem is the holy stillness of the environment fusing with the poet’s whole being.

Sound as Meaning 149 Table 5.5 Revision of the verbs Version

V1

V2

DEPTH

SHARPNESS

(1) (2) and (3) (4)

shimi shimi shimi

tsuku komu iru

+ +

+

There are two important lexical changes which contribute to a more effective representation of this fusion: the initial word and the verb of the poem. The ﬁrst word of the poem is changed from yamadera (‘mountain temple’) to sabishisa (‘loneliness’) and then to shizukasa (‘silence’). This revision implies a transition from the objective surroundings to a subjective feeling, and then to a fusion of the two, as the word shizukasa signiﬁes both the silence of the holy mountain temple and the tranquillity in the solitude of the poet. In other words, shizukasa captures a fusion of humans and nature, subject and object, a characteristic which underlies all Basho’s poetic works. The change of verbs from shimitsuku (‘seep and stick to’), to shimikomu (‘seep and include’) and then shimiiru (‘seep and pierce’) is also signiﬁcant to the notion of fusion. The three verbs used in each revision are compound verbs that display differences in depth and sharpness of penetration. Both these aspects are clearly seen to increase from versions (1) to (4) (Table 5.5), so that the verb, shimiiru, in the ﬁnal version most overtly enacts the fusion between the voice of the cicada and the silence of the rocks. The verbs are all used in a metaphorical context, as they describe the motion of the voice of the cicada as if it were liquid. This is apparent from the ﬁrst verb, shimu, which means that liquid seeps into or onto materials such as cloth, paper, soil or wood. Shimitsuku is a motion of superﬁcial seepage, shimikomu implies slightly deeper seepage, whereas shimiiru expresses both depth and sharpness of seepage. It is in this sense that shimiiru best depicts the force of the cicada’s voice as it meets, penetrates and is lost into the immense silence of the rocks. Phonological revisions The phonological changes affected by lexical revisions are even more compelling. The occurrence of ﬁve vowels, [i], [u], [e], [o] and [a], in the four versions indicates the dominance of [i] (Table 5.6).

150 Metaphor and Iconicity Table 5.6 Revision in the vowels Version

[i]

[u]

[e]

[o]

[a]

Total

(1) (2) (3) (4)

5 7 7 7

2 1 1 2

3 2 2 2

2 4 3 2

5 3 4 4

17 17 17 17

Table 5.7 Occurrence of vowels in the 50 haiku in Oku no Hosomichi

Occurrences %

[i]

[u]

[e]

[o]

[a]

Total

201 23.7

154 18.2

66 7.8

158 18.6

269 31.7

848 100

There is a slight increase in the occurrence of the vowel [i] from version (1) to versions (2), (3) and (4), compared to the occurrence of other vowels. What is especially interesting is that there is a rather high proportion of [i] in this haiku (7 out of 17 vowels, or 41.2 per cent), in fact about twice the average found for all 50 poems composed by Basho in his travel sketch, Oku no Hosonichi, in which this haiku is included (Table 5.7). The relative dominance of the vowel [i] is particularly important in this haiku because of its sound-symbolic effects. It is articulated with the least vocal energy among Japanese vowels, and hence symbolically suggests both short duration and small size. This sound-symbolism of [i] is mapped onto the image of the cicada, its small size and short lifespan (and possibly also onto human beings), in contrast to the eternal nature represented by the gigantic rocks surrounding the mountain temple. The sequential order of vowels also reveals an interesting structure within the poem (Table 5.8). The metric scheme forms a sound pattern from the ﬁrst to last line of LOW–HIGH–MID vowels, which is preserved from the ﬁrst version to the ﬁnal one. The dominance of [i] is located in the middle line, where the words iwa ni (‘into the rocks’) and shimi + verb (‘to seep + verb’ or ‘to penetrate’) appear. Again, the word shimiiru (‘to seep + pierce’) contributes to the dominance of [i]. Thus, the theme of this haiku – penetration into silence – is represented very clearly in its middle line by imagic iconicity of the vowel [i].

Sound as Meaning 151 Table 5.8 Sound pattern of vowels in sequence Metre Version

5 morae

(1) (2) (3) (4)

aa ai ai iu

e i i a

a a a a

a o a a

7 morae

5 morae

i i i i

e e e e

a a a a

i i i i

i i i i

i i i i

uu ou ou i u

i i i i

o o o o

o o o o

e e e e

Line 1

Line 2

Line 3

Dominance of LOW vowel [a]

Dominance of HIGH vowel [i]

Dominance of MID vowels [e] and [o]

In the process of revision, the consonants have undergone less change, perhaps because Basho’s original choice of consonant type already played an important sound-symbolic role in determining iconicity. The use of voiceless sibilant sounds such as [s] and [S], which typically symbolise silence in Japanese, as well as the dominance of voiceless obstruents (six out of seven obstruents), intentionally promotes the sound-symbolism of silence. There are only a few stops in the poem, and the voiceless [k] is the only one that survives into the ﬁnal version. Their place in version (4) of the poem seems to play a role, since one [k], three morae from the beginning of the poem, and the other, two from the end, contribute to bracketing or framing the entire poem. The former three versions do not have this kind of bracketing structure. Fusion of voice and silence It is clear from the discussion so far that the main theme of this haiku is the fusion of opposites: sound with silence. Through his intuitive insight, both Basho’s original and especially his revisions are aimed at this fusion. There is also a more holistic interaction of the sound and the meaning in the ﬁnal version. It would appear that in the ﬁnal version the semantic progression of the lines and the progression of their sounds go in opposite directions, as illustrated in Figure 5.3. The semantic progression of the lines goes from ‘silence’ to ‘voice’, namely from shizukasa ya (‘silence’) in Line 1 to semi no koe (‘voice of the cicada’) in Line 3. On the other hand, the phonological progression shows the reversed order ‘voice’ to ‘silence’.

152 Metaphor and Iconicity Line 3

SEMANTIC PROGRESSION

VOICE

Line 1 SILENCE

Line 1

PHONOLOGICAL PROGRESSION

Line 3

Figure 5.3 Phonological and semantic progressions of the text

Firstly, from Line 1 to Line 2, there is a repetition of the similar [-CONSONANTAL] sequence: Line 1 Line 2

S iz uk a i

w

a

Strictly speaking, [u] and [w] are different sounds; however, both have quite similar qualities, as they are articulated in the high, back position of the mouth. In this repetition, Line 2 totally lacks consonants as shown above. Secondly, in Lines 1 and 2, the consonants sandwiched by vowels lose their obstruency from the beginning of the line to its end: Line 1 Line 2

(k) a s a (n) i S i

aya imi

i

i (r)

And thirdly, from Lines 1 to 3, vowel sandwiches lose the Front–Back contrast between the sandwiching vowels (V1) and the sandwiched consonants (C):

Line Line Line Line

1 2a 2b 3

V1 [+BACK] [-BACK] [-BACK] [+BACK]

C [-BACK] [+BACK] [-BACK] [+BACK]

V1 [+BACK] [-BACK] [-BACK] [+BACK]

a s a, a y a iSi i m i, i i oko

In Lines 1 and 2a, vowels and consonants in the sandwich are articulated in contrastive positions in the mouth: vowels at the back and consonants at the front in Line 1; and vowels at the front and a consonant at the back in Line 2a; whereas in Lines 2b and 3, the places of articulation are non-contrastive between vowels and consonants.

Sound as Meaning 153

shared first mora

shi (zuk-) shi

mi (se)

mi

shared last mora

Figure 5.4 Oneness of silence and voice

All these facts imply that the sound shape of this haiku goes from ‘voice’ to ‘silence’. The hidden music of the three lines is a diagrammatic manifestation of the intended metaphorical meaning of the text, namely the fusion of voice and silence. The reversal of semantic and phonological progressions itself represents the fusion; there is no distinction between silence and voice nor between sound and meaning. There is perhaps a further observation that demonstrates the oneness of silence and voice.17 The three most important word/morphemes are shizuk-, shimi- and semi. Either the ﬁrst word (shizukasa ‘silence’) or the last (semi ‘cicada’) can be syntactically the subject of the middle one (shimiiru ‘to seep + pierce’). Also, shimi and shizuk- share the ﬁrst mora and shimi and semi share the last mora as shown graphically in Figure 5.4. Hence, shimi is like a phonetic compromise between shizuk- and semi. Finally, in Japanese onomatopoeia, the voice of cicadas is represented in four ways, and four types of cicada are named accordingly: [niinii] (or [minmin]), [Ji:Ji:], [kanakana] and [tsukutsuku bo:Si]. Notice that there are many [i], [a], [u], [m], [n] and [k] sounds in these onomatopoeic expressions. It is not known which kind of cicada Basho heard when he composed this haiku, but it is interesting to note that Basho used the same sounds as found in these onomatopoeic expressions of the cicada’s voices to compose this particular haiku about the cicada. Discussion In summary, the sounds and meanings of the poem as a whole, together help crystallise the moment of melting into the eternal stillness. The phonetic similarity motivates the semantic fusion of ‘silence’ and ‘voice’. At the same time, the semantic interpretation navigates diagrammatic reading of the text. As illustrated in Figure 5.5, the poem as global metaphor gets two major input spaces: ‘silence’ as a target, and ‘voice’ as a source, both being represented by the ﬁrst ﬁve-mora segment and the rest of the text, divided by kireji, ‘ya’. There is an emergent blended space, which fuses

154 Metaphor and Iconicity

Metaphorical Mapping

Iconic Mapping

Target

Source voice

Sound Shape

silence

onomatopoeia IMAGIC MAPPING sound-symbolism

IMAGIC MAPPING fusion DIAGRAMMATIC MAPPING IMAGIC MAPPING Blend MEANING

FORM

MORE CONTENT IS MORE FORM Grammatical Metaphors Figure 5.5 Metaphor–icon links in ‘Shizukasa ya’

‘voice’ and ‘silence’, and thereby leads the reader to the eternal stillness of nature that includes mountains, rocks, cicadas and the poet. All beings become silent – peaceful stillness. The sound shape of the poem is distinctively onomatopoeic and sound-symbolic. The onomatopoeia of the cicada’s cry has a reﬂection onto the various phonological segments used in the source. More importantly, however, the sound-symbolism of [i] for smallness and shortness, and [s] and [S] for silence is mapped onto the target input space and the blended space. The abundant use of these sound-symbolic segments is then interpreted diagrammatically by the grammatical metaphor of MORE CONTENT IS MORE FORM, to reinforce the interpretation of the text as a fusion of ‘sound’ and ‘voice’. This reading also navigates the diagrammatic reading of the sound structure such as the sequential order of sound segments, the role of voiceless obstruents and the phonological afﬁnity of morphemes. The detailed analysis of revisions, therefore, has

Sound as Meaning 155

made it clear that the ﬁnal version achieves the highest degree of interplay of metaphor and iconicity in terms of strength and complexity. Now, the question is whether Basho consciously used the principle of iconicity to revise his poem. And if so, how consciously? It is hard to resist speculating that the iconicity principle was working in the revising process to some extent as the analysis above has demonstrated. Jakobson (1985 [1980]: 69–70), being asked whether a conscious choice can be demonstrated through a comparison between different versions in certain writers, answers that there are three possibilities – chance, a subconscious activity and a conscious activity. Jakobson’s own preferred explanation lies with subconscious activity, although he admits that the notes, explanations and comments of the poets about their own compositions sometimes prove that they are quite conscious about their choices and their effects. I do not dare to measure how consciously Basho used iconicity by counting up the effects of his revisions. It is also beyond the scope of this study to judge whether or not his revisions were more oriented by semantic implications or by phonological effects. However, the strong phono-semantic link shown in his choice of words in the revising process – particularly the gap from the ﬁrst version to the others – seems to indicate rather clearly that this iconic link itself is crucial in achieving the effects of image and music aimed at in the poem. For example, there are two major phonological revisions provided by the lexical revision of the ﬁrst word from yamadera (‘mountain temple’) to sabishisa (‘loneliness’) and then to shizukasa (‘silence’). They are an increase of the vowel [i] and the sibilant consonants [s] and [S], both of which phono-symbolically suggest ‘silence’. This lexical revision must have been a conscious activity of the poet, and therefore, he might also have been quite conscious about the sound effects this revision would produce, because the iconic impact of the sounds is rather salient. The other phonological effects I have analysed in Basho’s revising process in support of the working of metaphor–icon links seem to be more subconsciously conceived by the poet, as they would only be detectable by this kind of microscopic dissection and examination.

Summary As the analysis of ‘The Bells’ and ‘Shizukasa ya’ has demonstrated, sound form in the poetic text plays an important role in conveying and creating meaning. Manifestations of imagic iconicity are immediate and strong, and prominent in the sample texts analysed above.

156 Metaphor and Iconicity

There may be few poems which depend on auditory imagic iconicity as heavily and as virtually as the poems looked at above. It seems that for the very reason that imagic iconicity is immediate and concrete, the vast majority of poets have avoided such an excessive and extensive use of onomatopoeia and sound-symbolism. This is most likely to be the case when the theme or the global metaphor of the text is irrelevant to the direct expressiveness of sounds. However, it is not difﬁcult to ﬁnd a poem which uses imagic iconicity in part, for a variety of different purposes. For example, poems often have a distinct line or passage with certain iconic sounds. Such a line or passage is made prominent by repetition or by position, to reinforce particular auditory qualities and, therefore, to differentiate the line or passage from the rest of the poetic text, and to suggest physical qualities or mental states by association of the sounds.

5.3 Diagrammatic structuring of the sound Generally speaking, it is diagrammatic iconicity that predominates over imagic iconicity in poetry. Sound patterns such as rhyme, metre and other formal structures contribute to the diagrammatic mapping between the sound form and the meaning it displays. For example, sound forms linked up by rhymed feet, assonance, alliteration or other phonological or syntactic parallelism are regarded as equivalent, and cue the semantic equivalence between them in terms of the grammatical metaphor SIMILARITY IN CONTENT IS SIMILARITY IN FORM. It is the structuring of sounds in addition to their inherent qualities that matters in diagrammatic iconicity in poetry. Hence, manifestations of diagrammatic iconicity are germane to a more sophisticated and perhaps more subtle understanding of the text. Works by Jakobson and his advocates (Jakobson 1960, 1971, Shapiro 1976, 1998, Freeman 1978a, b, Jakobson and Waugh 1979, Ross 1981, 1982a, b, 1986, 1989, 1990, 2000, among others) contribute to clarifying auditory and syntactic manifestations of iconicity in English, French, Spanish and Russian poetic texts. Although their works (except for Freeman 1978a, b, Shapiro 1998) have a focus primarily on the formal analysis of structure rather than the manifestation of iconicity per se, what has prevailed in their structural analyses is the diagrammatic structuring of sounds and syntax in reinforcing or enacting the semantic content of the text. They offer manifold examples in which rhyme, alliteration and metrical parallelism manifest diagrammatic iconicity to the meaning they express. Against the backdrop of such research tradition, this section analyses

Sound as Meaning 157

two illustrations of subtler diagrammatic manifestations, in order to claim that even simpler conﬁgurations of sounds than rhyme and metre can achieve powerful iconic structuring. In effect, I propose that the principle of abstract diagrammatic iconicity is an effective methodological tool for interpreting the sound structure of texts to be looked at. The ﬁrst example is an English free verse, ‘The Red Wheelbarrow’, by William Carlos Williams. The second example is a traditional tanka (31mora poem) by Tomonori Kino. Both examples are very short in form, encapsulating a simple impression of a natural scene. At ﬁrst glance, neither of the two poems is particularly iconic. Their diagrammatic iconicity is not readily recognisable. It is only through detailed linguistic analysis that certain conﬁgurations of sounds receive iconic interpretations consonant with the metaphorical meaning of the text. This is what the following section is intended to show. 5.3.1 Mirror of sounds in Williams’s ‘The Red Wheelbarrow’ Example 5.4 (Williams 1970 [1923]: 138) The Red Wheelbarrow So much depends upon

1 2

a red wheel barrow

3 4

glazed with rain water

5 6

beside the white chickens.

7 8

This particular poem has been chosen for the following two reasons: that brevity and simplicity of its representation are comparable to, and perhaps are inﬂuenced by, haiku;18 and that it does not exhibit the traditional device of English rhyme or metre. Ross (1982c, 1986: 187–95) presents an in-depth analysis of the relationship of meaning to form of this poem, and demonstrates remarkable structuring of sound and syntax, although he does not speciﬁcally develop his argument on the concept of metaphor or iconicity. Here I would like to reinterpret his claims in my framework of the interplay between metaphor and iconicity, because the form–meaning correspondences that his analysis has clariﬁed provide an optimal example of diagrammatic iconicity.

158 Metaphor and Iconicity

The tableau of the poem The poem depicts the ‘tableau’ of the wheelbarrow, rainwater and chickens, or the simple natural scene of a farmer’s yard after the rain. Williams uses three colours of pigments – red, transparent and white. He selects three basic objects – artefact, water and animal. It is as a snapshot rather than a motion picture that these objects are displayed. There are two grammatical factors that support the interpretation of the poem as a snapshot. Syntactically, the way Williams describes the location of the poem’s two concrete nouns, the wheelbarrow and the chickens, deserves attention. The normal way of describing the relationship of movable and immovable objects in English is to highlight or foreground the movable and to put the immovable in the background. This is usually done by the use of such sentence constructions as ‘X (movable) is Preposition Y (immovable)’, or ‘there is X (movable) Preposition Y (immovable)’. Hence, it is more natural to say, ‘the chickens (movable) are beside the wheelbarrow (immovable)’ than ‘the wheelbarrow is beside the chickens’.19 The poem reverses this normal order by placing the wheelbarrow beside the chickens; and hence, makes one see the chickens as if their motion is stopped by the grammatical metaphor, IMMOVABLE IS BACKGROUND. This syntactic implication of the chickens as an immovable entity contributes to the ‘tableau’ reading of the poem. Secondly, the other concrete noun of the poem, the wheelbarrow, also receives a static interpretation. The wheelbarrow can be a movable object only when it is used by a human agent. In this poem, however, there is no overt mention of any human agents. Moreover, the verb ‘glaze’, which modiﬁes the wheelbarrow, is etymologically derived from a noun ‘glass’. The rainwater which has washed the dirt away and wets the surface of the wheelbarrow makes it a looking glass. As Ross (1986: 188) puts it, ‘Williams has made this a static scene: by calling the water a glaze, he has stopped it from ﬂowing, made it into a mirror.’ The mirror-image The metaphor of a mirror is indeed a navigating force of auditory diagrammatic manifestations. Both syntactically, semantically and geometrically, the verb ‘glaze’ is a pivot word of the poem. It is a syntactic turning point where the change from subordination to coordination occurs. It is a semantic turning point where the metaphor of a mirror emerges by the verb’s etymological base, ‘glass’. The word is geometrically located at the centre of the poem.

Sound as Meaning 159

Moreover, as Ross (1986) discloses in his detailed analysis, it is at this pivotal word that the sound structure of the text functions as a mirror. The ﬁrst two stanzas and the last two stanzas form a phonological mirror-image in the following respects. Firstly, the poem’s four stanzas consist of four words each: three on the ﬁrst line and one on the second line. In terms of the number of syllables (and so, of the vowels), the ﬁrst and the last stanzas consist of six syllables, whereas the second and the third have ﬁve syllables. I II III IV

so/ much/ de/ pends/ up/ on a/ red/ wheel/ ba/ rrow grazed/ with/ rain/ wa/ ter be/ side/ the/ white/ chi/ kins

(6) (5) (5) (6)

Thus, the number of syllables in each stanza forms a mirror-image as shown in Figure 5.6. Secondly, out of the poem’s 22 vowels, the distribution of the front, mid and back vowels supports the mirroring pattern that sets peripheral stanzas, I and IV, off against central ones, II and III, as indicated in Table 5.9. Thirdly, almost all of the consonants occur in pairs.20 Particularly, [d], [n] and [r] occur four times, and manifest the mirror structure as well (see Table 5.10). [d] occurs in even lines in each stanza in a mirrorimage sequence of disyllable–monosyllable–monosyllable–disyllable. Three out of four [n]’s occur in line-ﬁnal positions and in a mirrorimage sequence of [-nz]–[-n]–[-n]–[-nz], and disyllable–monosyllable– monosyllable–disyllable. [r] seems to connect the middle stanzas in an alternate order between monosyllables and disyllables, and [r]’s occurrence forms a mirror-image sequence of 0–2–2–0.

I

so much depends upon

II

a red wheel barrow

III

glazed with rain water

IV

beside the white chickens.

Figure 5.6 Mirror-image of syllables

160 Metaphor and Iconicity Table 5.9 Mirror-image of vowels Vowels Stanza

Front

Mid

Back

I

1 ([e])

3 ([V][R][R])

䉳 2 ([ou][A])

II

3 ([e][i:][æ])

1 ([R])

1 ([ou])

䉳

III

3 ([ei][i][ei])

1 ([R])

1 ([O:])

䉳

IV

1 ([i])

3 ([R] [R][R])

2 䉳 ([ai][ai])

Table 5.10 Mirror-image of consonants Stanza

[d]

[n]

[r]

I II III IV

depends Red glazed beside

depends, upon 0 rain chickens

0 red, barrow rain, water 0

Figure 5.7 summarises in a graphic representation the interplay of metaphor and iconicity in ‘The Red Wheelbarrow’. The ‘tableau’ of the wheelbarrow, rainwater and chickens constitutes a source input space. Just like the juxtaposition of haiku, the poem’s subject, ‘so much’, as a target input space, is juxtaposed or equated with this static scene subordinated by the verb phrase, ‘depends upon’. Williams then leaves the reader to infer what is meant by ‘so much’. These two input spaces yield an emergent blended space, in which the wheelbarrow can be a mirror physically reﬂecting the fresh scene of the farm washed by the rain, and at the same time, a mirror reﬂecting the mental world of the farmer’s life. It can be this mirror of physical and mental reﬂection that ‘so much’ depends upon. The reading of the poem as a metaphor of mirror reﬂecting the farmer’s mental and physical worlds navigates another level of interpretation. It leads on to suggest a diagrammatic iconic mapping from the blended space onto the form as indicated by the arrow in Figure 5.7.

Sound as Meaning 161

Metaphorical Mapping

Iconic Mapping

Sound Shape

Source

Target

tableau

so much

wheelbarrow rainwater mirror chickens DIAGRAMMATIC MAPPING

mirror reflection

Blend

FORM

MEANING

MIRROR Grammatical Metaphors Figure 5.7 Metaphor–icon links in ‘The Red Wheelbarrow’

The diagrammatic structuring of sounds in the text becomes apparent as a result of this metaphorical reading. The layers of mirror-image sequence of sounds, as analysed above, are charged by the metaphorical meaning of the text. Once this interplay is manifested, then the formal auditory features reinforce the semantic interpretation in turn. 5.3.2 Hidden melodies in Tomonori’s ‘Hisakata no’ The second example is taken from a completely different poetic and linguistic tradition of a completely different historical era; but it displays similar diagrammatic manifestations.21 The poem for analysis is a famous tanka composed by Tomonori Kino approximately 1000 years ago, and compiled in Kokin Waka-Shu (Ozawa 1971 [905]),22 the second oldest anthology of poetry in Japan. It is very short in form, encapsulating a simple impression of a natural scene. At

162 Metaphor and Iconicity

ﬁrst glance, the poem is not particularly iconic. Its diagrammatic iconicity is not readily recognisable. It is only through a detailed linguistic analysis that certain conﬁgurations of sounds receive iconic interpretations consonant with the metaphorical meaning of the text. The following gives kana and phonetic transcriptions of the poem and its word-for-word English translation. Example 5.5 (Tomonori Kino 1971 [905]: 88–9)

1 2 3 4 5

phonetic notation hisakata no23 hikari nodokeki haru no hi ni Sizukokoro naku hana no tSiruramu

word-for-word translation distance GEN light gentle spring of day on quiet.mind/heart there.not.be ﬂower SUB fall.AUX

‘On a day of a spring whose light is gentle, I wonder why without having a quiet mind/heart, the cherry blossoms are falling’ (my translation). Heart of ﬂowers The poem depicts falling cherry blossoms on a tranquil spring day. Although the sun is shining, the petals are showering down without being able to ﬁnd any peace of mind. The ﬁrst three lines describe the time, which is a calm spring day full of sunlight. They provide the background setting for an event that is described in the last line – the falling of cherry blossoms. Hence, Lines 1, 2, 3 and 5 basically describe the external reality. Line 4, however, is different. It contains the poem’s only metaphorical word, shizukokoro (‘quiet.mind/heart’). Japanese does not differentiate mind and heart in its vocabulary. Kokoro designates this undifferentiated entity of mind and heart. Here, the ﬂowers are personiﬁed as if they had kokoro. At the same time, humans including the poet himself are ‘plantised’, as if the sorrow of the falling ﬂowers became one with the sorrow of the human viewers. Indeed, shizukokoro is a semantic turning point, from which the calm and smooth setting has suddenly become problematic – lack of peace in kokoro and the falling of ﬂowers. Line 4 is the only line with a negative word, naku (‘without’). The last line has a full description of the event, the falling of cherry blossoms. The last morpheme of the poem, -ramu (‘I wonder why’), which is a modal auxiliary, concludes as well as enacts the fusion or identiﬁcation of a subjective feeling

Sound as Meaning 163

of sorrow felt by the poet and an objective event of falling cherry petals. Consequently, one interpretation could be that the poet has a melancholy feeling about the transience of the beauty of cherry blossoms. His heart, feeling the sorrow of the ﬂowers, becomes one with the hearts of the ﬂowers. In what follows, I will show that the poem displays strong iconic correspondences between the phonological and the semantic aspects. When analysed in a similar way to the previous sections, the text exhibits three aspects which are phonologically iconic to its interpretation: the distinctiveness of Line 4; the distribution of certain repeatedly used consonants, [h], [n], [k] and [r]; and the distribution of two vowels, [a] and [o]. Distinctiveness of Line 4 There are a number of phonological aspects in which Line 4 is different. Firstly, it is different from other lines in terms of the initial sound, which goes like h-h-h- S -h. Secondly, all nouns in the poem start with [h] except for kokoro in Line 4. Thirdly, Line 4 has the only compound noun of the poem, shizukokoro (‘quiet.mind/heart’), which is followed by the mora [na], while in other lines, the ﬁrst nouns, hisakata (‘distance’), hikari (‘light’), haru (‘spring’) and hana (‘ﬂower’) are all followed by the mora [no]. Fourthly, in Lines 1, 2, 3 and 5, the sequential order of the ﬁrst three consonants has a regular progressive pattern. The pattern is that the third consonant in the former line becomes the second consonant in the next line. Notice that the consonant [k] appears in the third mora in Line 1 and in the second in Line 2; [r] appears in the third mora in Line 2 and in the second in Line 3; and [n] appears in the third mora in Line 3 and in the second in Line 5. Line 4 has no such pattern. All these differences of Line 4 can be interpreted as enacting with the semantic content of the line that there is no calm kokoro for the falling cherry blossoms or for the poet himself. The meaning of the poem navigates the reading of the phonological structural patterns in such a way that they resemble the meaning they represent. Hidden melodies of hana and kokoro As Table 5.11 illustrates, there are four consonants that are used repeatedly in the text: 5 [h]’s, 7 [n]’s, 7 [k]’s and 5 [r]’s. Other consonants occur only once in the text.

164 Metaphor and Iconicity Table 5.11 Number of occurrence of consonants Obstruents

Sonorants

Line

t

d

s

z

Ú

tÚ

k

m

n

r

h

1 2 3 4 5

1 0 0 0 0

0 1 0 0 0

1 0 0 0 0

0 0 0 1 0

0 0 0 1 0

0 0 0 0 1

1 3 0 3 0

0 0 0 0 1

1 1 2 1 2

0 1 1 1 2

1 1 2 0 1

Sub-total

1

1

1

1

1

1

7

1

7

5

5

Total

13

1

hisakata no

2

hikari nodokeki

3

haru no hi ni

4

∫izukokoro naku

5

hana no t∫iruramu

18

Figure 5.8 Distribution of [h] and [n]

The overwhelming repetition of these four consonants shows an interesting distribution, which needs some explanation. Figure 5.8 indicates a distribution of [h] and [n]. The distribution of [h] and [n] follows an alternating pattern until Line 4, where the pattern stops because of the lack of [h]. In Line 5, the pattern revives in hana, but then vanishes in ‘no tSiruramu’. The distance between [h] and [n] in each [h]–[n] pair becomes closer and closer, as there are three morae between [h] and [n] in ‘hisakata no’, two morae in ‘hikari nodokeki’, one mora in ‘haru no’, zero mora and one word boundary in ‘hi ni’, and ﬁnally no mora or word boundary in ‘hana’. From this, it is speculated that [h] and [n] are a hidden24 preﬁguring of the hana (‘ﬂowers’ or ‘cherry blossoms’). This melody can be heard from the very beginning of the poem. It gets clearer and clearer as one proceeds towards hana in Line 5 (with the sole exception of Line 4), and then cannot be heard because the ﬂowers are falling, and the dropping of the petals dissolves the h–n melody as well.

Sound as Meaning 165 1

hisakata no

2

hikari nodokeki

3

haru no hi ni

4

∫izukokoro naku

5

hana no t∫iruramu

Figure 5.9 Distribution of [k] and [r]

The distribution of [k] and [r] displays a remarkably similar conﬁguration. As shown in Figure 5.9, [k] and [r] form a pattern of k–k–r, which is a hidden preﬁguring of the kokoro (‘mind/heart’). The distance between each sound of the k–k–r trios also gets closer and closer as the lines progress until the word kokoro, in which there is no mora or word boundary between [k], [k] and [r]. The k–k–r melody keeps playing until kokoro in Line 4, and then dissolves, because there can be no melody when there is no calm kokoro. These phonological structures can be seen as iconic manifestations of the semantic content, navigated by the grammatical metaphors such as IMPORTANCE IS ABUNDANCE, PATTERN IN MEANING IS PATTERN IN FORM and SIMILARITY IN MEANING IS SIMILARITY IN FORM. Hence, it is summarised that: 1. Hana and kokoro are the two keywords in this poetic text; 2. The two melodies show again a strong diagrammatic iconicity to the semantic architecture of the poem – the decay of the hana melody represents the falling of the petals of the hana; whereas the decay of kokoro melody represents the absence of peace of kokoro. In addition, since the decay of both of the hidden melodies begins in Line 4, this fact also joins the two words; 3. The fact that hana and kokoro both participate in similar melodic decays and that their decays play similar iconic roles in the semiotic architecture of the poem would suggest that they are poetically equivalent. Thus, it is a plausible interpretation that hana and kokoro fuse with each other in the harmony of these two hidden melodies – that hana is kokoro (the mind/heart of human being and the mind/heart of cherry blossoms is one). Again, the stage for this fusion is set by the radical distinctiveness of Line 4.

166 Metaphor and Iconicity

Kokoro inside hana It has been argued that one vantage point from which to view the poem is the one which sees how different the fourth line is from the other four lines, and that there is a precise reason for its deviance. In Line 4, just as the h–n melody of hana is interrupted, the k–k–r melody of kokoro consolidates to reveal its secret. Therefore, in a sense, kokoro’s melody is ‘inside’ hana’s melody – the ﬁrst [h] is before the ﬁrst [k], and k–k– r-’s blossoming in Line 4, is before h–n’s blossoming in Line 5. The following discussion shows some patterns that put kokoro ‘inside’ hana in a different but complementary way. The pattern that emerges is one which connects the ﬁrst and last lines, and also the second and fourth lines, surrounding, in an ‘onion-skin’ fashion, the pivotal line, Line 3. The ﬁrst thing to notice, in support of this ‘onion-skin’ pattern, is the syntactic fact that there are just two adjectival notions in the poem – nodokeki (‘gentle’) in Line 2 and shizu(‘quiet’) in Line 4 – all in interior lines. The poem’s part-of-speech structure indicates that only the third pivotal line has two nouns, haru and hi, while every other line has one. The nouns in the exterior lines are followed by the particle no (although engaging in different grammatical functions – no in Line 1 is a genitive marker, while no in Line 5 is a subject marker), whereas those in the even, adjectival lines are the poem’s only bare nouns. Concerning the phonological properties, there is support for postulating an onion-skin pattern as well. One of the primary contrasts between the two focal words, kokoro and hana, is in vowel height – mid and low. It is noted that the last line’s three [a]’s are matched exactly by three [a]’s in the ﬁrst line, with each of the interior lines having just one. Just as hana is characterised by a sequence of two adjacent low vowels, and one ﬁrst hears low vowels in the three adjacent morae of Line 1’s [(hi)sakata], so the characteristic mid vowels of kokoro are also introduced by a sequence of three mid vowels in Line 2’s [nodoke(ki)]. The odd lines manifest exactly one mid vowel apiece. Consequently, one can conclude, from these syntactic and phonological facts, that exterior lines and interior lines are poetically equivalent, and that the former represent hana, while the latter represent kokoro. Hence, the onion-skin structure of ‘kokoro inside hana’ in Figure 5.10 could also be interpreted as a fusion of hana and kokoro. This is, at the same time, a fusion between nature and humanity – one of the classical themes in Japanese poetic arts.

Sound as Meaning 167

Figure 5.10

1

hisakata no

2

hikari nodokeki

3

haru no hi ni

4

∫izukokoro naku

5

hana no t∫iruramu

hana

kokoro

Kokoro inside hana

Discussion The poem as a global metaphor comprises two input spaces: NATURE as a source and HUMAN as a target. NATURE is prototypically represented by cherry blossoms, which are falling away in spite of the calm spring day. HUMAN is represented by the poet (or humanity in general) with melancholic kokoro over the transience of beauty. In the blended space, there is an emergent structure of fusion of NATURE and HUMAN, as both cherry blossoms and human beings have kokoro. The poem as a global metaphor describes human feeling in terms of kokoro of cherry blossoms. As illustrated in Figure 5.11, this metaphorical reading then navigates an iconic interpretation of several abstract patterns in the distribution of linguistic elements in the text. Particularly, there is a strong diagrammatic mapping between form and meaning in the following three main ways: 1. The phonological distinctiveness of Line 4 is consonant with the meaning of this line; namely, a lack of orderly sound patterns represents a lack of quiet kokoro. 2. The two vital words, hana and kokoro, inspire hidden melodies, and the progression and dissolution of these melodies are iconic to the semantics of the poem – falling of cherry blossoms is represented by the decay of the hana melody, the absence of peace in mind/heart by the dissolution of the kokoro melody. In addition, a fusion of hana and kokoro is represented by the equivalent roles these hidden melodies play in the whole text. 3. The parallelism of these two melodies as well as the syntactic and phonological equivalences shown in the onion-skin structure of exterior and interior lines also suggests that the fusion of human kokoro with the ﬂower is iconically embodied in the very structure of this poem.

168 Metaphor and Iconicity

Metaphorical Mapping

Iconic Mapping

Sound Patterns Line 4 Hidden Melodies Onion-Skin

Source

Target

nature

human

spring light hana kokoro kokoro DIAGRAMMATIC MAPPING

fusion hana kokoro Blend

FORM

MEANING

SIMILARITY IN CONTENT IS SIMILARITY IN FORM IMPORTANCE IS ABUNDANCE Grammatical Metaphors Figure 5.11 Metaphor–icon links in ‘Hisakata no’

5.4 Concluding remarks This chapter has looked at auditory manifestations of the interplay between metaphor and iconicity in several texts. The major claim has been that despite the differences in the phonological systems of English and Japanese, iconicity is an important motivation in the construction and interpretation of texts examined. To put it differently, the analyses have shown that the sound form not only states the meaning but also enacts it in a variety of ways. Japanese and English differ in the inventory and the structure of sound materials that poets can use as their resources for creation. English has a greater number of phonemes, in both consonants and vowels, and a more complex make-up of syllables than Japanese. The construction of English syllables allows consonant clusters, diphthongs

Sound as Meaning 169

and long vowels, whereas Japanese morae have rather simple constructions such as [CV] and [V]. With a more varied inventory and more complex sound structures, the English literary tradition has evolved more sophisticated phonological devices for poetry, such as various forms of rhyme and metre. In contrast, Japanese has far simpler sound structures than English. The Japanese tradition has developed distinctive poetic forms, haiku and tanka, which are characterised by the simple moraic metre – an alternation of ﬁve- and seven-mora segments. In addition, the abundance of literature on English prosody or metrics reinforces the common conception that sound is considered essential to poetry. Such a conception is expressed, for example, by Jakobson and Waugh (1979: 215): ‘the notion of verse implies the indispensable presence of a certain speciﬁc, ad hoc organization of the verbal sound matter’. By contrast, the phonological simplicity of traditional poetic forms may therefore explain the fact that there have been very few works on the phonological aspect of Japanese poetry.25 To conclude, despite the fact that English and Japanese have different sound systems and structures, and hence have evolved different poetic traditions, both exhibit auditory iconicity. Manifestations of imagic and diagrammatic iconicity may vary from one text to another; but this chapter has demonstrated that the cognitive perspective provides an effective methodology towards a more comprehensive analysis of the dynamism of meaning creation through the interplay between metaphor and iconicity.

This page intentionally left blank

Part III Further Issues

This page intentionally left blank

6 The Interplay of Metaphor and Iconicity in Grammar and Discourse

Syntax is not independent of meaning, especially metaphorical aspects of meaning. George Lakoff and Mark Johnson (1980) The preceding chapters (Chapters 1–5) deal with the interplay of metaphor and iconicity in the foregrounded case of poetic texts. By focusing on a number of short Japanese and English poems, these chapters were devoted to the demonstration of metaphor–icon links and their complexity, both at the visual and the auditory levels, in their two basic manifestations: iconic moments residing in metaphor; and metaphor giving iconic interpretation to form. Throughout the analyses it has been argued that the relationship between form and meaning in poetic texts is primarily motivated rather than arbitrary. This chapter takes up the same argument in the case of more general linguistic issues. A special focus will be put on the discussion of grammatical metaphors, on the one hand, and the issue of polite language, on the other.

6.1 A typology of grammatical metaphors The interplay between metaphor and iconicity in general linguistic resources (such as sounds, words, sentences, discourses and writing systems) can be detected in the way in which a certain linguistic form receives an iconic interpretation by metaphors governing the conception of language. Lakoff and Johnson (1980: 126–38) note there are conceptual metaphors in grammar that can provide direct links between form and content. I termed these metaphors ‘grammatical metaphors’ 173

174 Metaphor and Iconicity

and offered a general overview of the functions of these metaphors in Chapter 2. To elaborate the major points discussed in Chapter 2, this section attempts a more systematic classiﬁcation of grammatical metaphors which concern the relationship of form and content in grammatical conventions in phonology, morphology, word formation, word order and pragmatic usage. Basic grammatical metaphors: the

CONDUIT

metaphor

As discussed before, the CONDUIT metaphor (Reddy 1979) is pervasive in the conceptualisation of language and communication crosslinguistically in literate societies. Ideas are conceptualised as objects, linguistic expressions as containers and communication as sending and receiving the containers containing ideas as objects. The CONDUIT is a linear entity and, therefore, language and communication, speaking and writing, are conceptualised in terms of space in a linear order. The CONDUIT metaphor entails some basic grammatical metaphors in terms of an entity and in terms of space: LINGUISTIC FORM IS AN ENTITY, CONTENT (MEANING) DEPENDS ON LOCATION OF FORM and CONTENT (MEANING) DEPENDS ON SEQUENCE OF FORM. Linguistic forms, such as a sound, a word, a phrase, a sentence and a discourse, are treated as if they were a separable and delineated entity. In real speech, it is difﬁcult to delineate these forms separately because of the continuous nature of the sound ﬂow. The units employed in the writing systems substantiate the existence of the LINGUISTIC FORM IS AN ENTITY metaphor. For example, European languages write words separately from each other with a blank space in between, because words are conceptualised as entities. Japanese kana syllabaries, on the other hand, correspond with a moraic unit, which is conceptualised as an entity. Punctuation also supports the LINGUISTIC FORM IS AN ENTITY metaphor. A full stop symbol, for example, not only signals the end of a sentence but also delineates the sentence as a distinctive entity. Likewise, quotation marks delineates what is quoted as an entity. The location and sequential order of the linguistic units also relate to the interpretation of the meaning expressed. For example, when two words are put closer in location, the meaning might be affected by the location as shown in Example 6.1. Example 6.1 (Posner 1986: 319) (a) the uncompleted fourth symphony (b) the fourth uncompleted symphony

The Interplay of Metaphor and Iconicity in Grammar and Discourse 175

Example 6.1(a) means that the fourth symphony is uncompleted, whereas Example 6.1(b) means that it is the fourth of the uncompleted symphonies. The difference of interpretation indicates that the attribute standing nearest to the head noun forms a semantic unit with it. Likewise, the sequence of entities differentiates the meaning as in Example 6.2. Example 6.2 (a) Tom came back home and studied. (b) Tom studied and came back home. The difference of interpretation between Examples 6.2(a) and (b) depends on the sequential order of verbs with the use of ‘and’. In Example 6.2(a), Tom came back home before he studied because that is the sequential order of mention. In Example 6.2(b), the order is reversed, and so is the content. Elaborated grammatical metaphors The basic grammatical metaphors, concerning the ontology and the orientation of linguistic forms, are further elaborated and combined to form a handful of grammatical metaphors. The manifestations of these elaborated grammatical metaphors can be roughly divided into two types: relational mapping and structural mapping. These correspond with what I termed in Chapter 2 ‘relational diagram’ and ‘structural diagram’ because the mapping between form and meaning in both manifestations is diagrammatic in nature. ‘Relational diagrams’ display a correspondence of the similarity in form with the similarity in meaning, and of the difference in form with the difference in meaning as illustrated in Figure 6.1 (previously mentioned as Figure 2.4). ‘Structural diagrams’, on the other hand, display a correspondence between structure of form and structure of content as shown in Figure 6.2 (previously mentioned as Figure 2.3). 6.1.1 Relational diagrams The grammatical metaphors of relational diagrams can be stated in the following way: SIMILARITY IN MEANING IS SIMILARITY IN FORM; DIFFERENCE IN MEANING IS DIFFERENCE IN FORM. As indicated above, the isomorphic principle has two components: difference in meaning is reﬂected in the difference in form; and similarity in meaning is reﬂected in the similarity in form.

176 Metaphor and Iconicity

Iconic Mapping

DIFFERENCE IN MEANING IS DIFFERENCE IN FORM

SIMILARITY IN MEANING IS SIMILARITY IN FORM

FORM

MEANING

Figure 6.1 Relational diagram

Difference of form Difference in form implies a difference in meaning, but it does not necessarily inform the nature or degree of difference. When two words or sentences have different forms, they mean different things, however subtle that change in meaning may be. Although this contention has been challenged by the claim that there are synonyms and paraphrases in language, different forms that mean the same thing, Bolinger (1977) has demonstrated quite convincingly that ‘there is no difference in form without some difference in meaning’ (Bolinger 1977: vi), using a number of examples from English and Spanish. Haiman (1985a: 21) also argues that total synonymy is rare in language.

The Interplay of Metaphor and Iconicity in Grammar and Discourse 177

To illustrate the point, I would like to consider a few examples taken from Bolinger (1977: 10–14). It is generally argued that the gerund and the inﬁnitive are synonymous, as in Example 6.3. Example 6.3 (a) Waiting would have been a mistake. (b) To wait would have been a mistake. However, the ungrammaticality of Example 6.4(b) conﬂicts with this claim. Example 6.4 (a) Waiting has been a mistake. (b) *To wait has been a mistake. The conﬂict comes from the ‘hypothetical’ meaning attached to the inﬁnitive. Therefore, the gerund and the inﬁnitive are not always synonymous, because they are different in form. Likewise, it is commonly claimed that the active voice and the passive voice are synonymous, as in Example 6.5. Example 6.5 (a) George turned the pages. (b) The pages were turned by George. However, Example 6.6(b) is ungrammatical. Example 6.6 (a) George turned the corner. (b) *The corner was turned by George. The passive voice is used when the effect is produced on the patient. In Example 6.5, something happens to the pages in the process, whereas in Example 6.6, the effect cannot be produced on the patient because the patient is the ‘corner’. Similarity of form Similarity in form signals similarity in meaning. Waugh and Newﬁeld (1995) demonstrate that the lexicon is pervaded by a host of associations between words based on form–meaning connections, which then result in word–afﬁnity relations. For example, an initial /ﬂ–/ (phonestheme) in English is expressive of movement and characterises a whole family of words.

178 Metaphor and Iconicity

Example 6.7 ﬂap, ﬂare, ﬂee, ﬂick, ﬂicker, ﬂing, ﬂip, ﬂit, ﬂitter, ﬂow, ﬂutter, ﬂy In a similar way, the word families of demonstratives and interrogatives indicate an isomorphic association of meaning cued by similarity of form. Example 6.8 the, this, that, they, their, thee, thou, thy, thine, then, there, thus, than, though (Bloomﬁeld 1933: 47, 244) Example 6.9 what, why, when, where, which, whether, how (Jakobson and Waugh 1979: 55) Waugh and Newﬁeld (1995) further argue that many such word afﬁnity phenomena are best understood as being based on a paradigmatic network of correspondences between form and meaning in words as a gestalt (uniﬁed whole), rather than on a syntagmatic analysis of correspondence between form and meaning in segmentable morphemes. Nonetheless, they present clear cases of grammatical metaphor based on the relational diagrammatic mapping: SIMILAR MEANING IS SIMILAR FORM. 6.1.2 Structural diagrams There are several grammatical metaphors which navigate the interpretation of form on the basis of the structural analogy between form and meaning. In other words, these metaphors reinforce the fact that a certain structure of linguistic representation, such as linearity, proximity, quantity, symmetry and location, is motivated by its similarity to the content structure it represents. A brief non-exhaustive list of such metaphors include TEMPORALITY IS LINEARITY; ‘ME FIRST’; RELEVANCE IS CLOSENESS; MORE CONTENT IS MORE FORM; SYMMETRICAL MEANING IS SYMMETRICAL FORM; ASYMMETRICAL MEANING IS ASYMMETRICAL FORM; MARKED MEANING IS MARKED FORM; TIME IS PLACE; CAUSE IS PLACE and DEFERENCE IS DISTANCE. TEMPORALITY IS LINEARITY The linear order of mention corresponds to the temporal sequence of concepts mentioned. When describing a series of actions occurring in time, the normal sentence reports them in the same order as they occur in reality.

The Interplay of Metaphor and Iconicity in Grammar and Discourse 179

Iconic Mapping

SEQUENCE IN TIME IS SEQUENCE IN FORM

QUANTITY OF CONTENT IS QUANTITY OF FORM FORM

MEANING

Figure 6.2 Structural diagram

Example 6.10 (a) Mary came in and sat down. and not (b) ?Mary sat down and came in. The conjunction ‘and’ in these sentences has an asymmetric use,1 which clearly promotes the iconic conventions of narrative word order.2 The grammatical metaphor, TEMPORALITY IS LINEARITY, is arguably more important in a non-inﬂectional language such as Chinese, because temporal sequence is expressed by word order. Tai (1985: 50), for example, claims in relation to Chinese that ‘the relative word order between two syntactic units is determined by the temporal order of the states which they represent in the conceptual world’. ‘ME FIRST’ (Cooper and Ross 1975) A kind of linear iconicity also shows up in the set of frozen expressions of two antithetical conjuncts such as ‘up and down’, ‘now and then’

180 Metaphor and Iconicity

and ‘men and women’. The reverse order does not hold: ‘*down and up’, ‘*then and now’ or ‘*women and men’. Cooper and Ross (1975) demonstrate that in such frozen expressions the properties of the prototypical person, i.e. what they call ME (whose properties are UP, FRONT, ACTIVE, GOOD, HERE, NOW, etc.), come ﬁrst. Japanese frozen expressions, in which two conjuncts are put side by side without any conjunction such as ‘and’, also indicate a similar linear iconicity supported by the ME FIRST metaphor, in which the unmarked comes ﬁrst, the marked second (Fujii 1986) as in Example 6.11. Example 6.11 (a) ue-shita up-down ‘up and down’ (b) *shita ue (c) danjo men.women ‘men and women’ (d) *jodan SEMANTIC RELEVANCE IS CLOSENESS Not only linearity but also proximity of word order suggests iconic interpretation, i.e. elements that occur closer together tend to be semantically closer. For example, Bybee (1985) demonstrates that there is a universal tendency in which the closeness between verb stems and inﬂectional categories corresponds to the conceptual proximity of the meaning of a verb and the concepts these categories express. The inﬂectional categories examined in her survey of 50 languages of the world are valence, voice, aspect, tense, mood, agreement of number, of person and of gender. Her results show that: (1) The more relevant a category is to the verb, the more likely it is to occur in a synthetic or bound construction with the verb: (2) The more relevant a morphological category is to the verb, the closer its marker will occur with respect to the verb stem: (3) The more relevant a morphological category is to the verb, the greater will be the morpho-phonological fusion of that category with the stem. (Bybee 1985: 11–12) A further interesting manifestation of proximity iconicity was noted by Posner (1986) in the order of attributive adjectives to the head noun

The Interplay of Metaphor and Iconicity in Grammar and Discourse 181

in English and German. He points out two types of proximity iconicity: ‘Local proximity to the head noun indicates nouniness’ (Posner 1986: 316);3 and ‘Local proximity indicates semantic closeness’ (Posner 1986: 320). Lakoff and Johnson (1980: 126–33) discuss similar phenomena from the perspective of a cognitive metaphor, STRENGTH OF EFFECT IS CLOSENESS. Compare the two similar sentences below. Example 6.12 (a) Mary doesn’t think he’ll leave until tomorrow. (b) Mary thinks he won’t leave until tomorrow. It is observed that Example 6.12 (a) has a weaker negative force than (b) as the negative marker ‘n’t’ is further away from the verb ‘leave’. Likewise in the following pairs of sentences, the proximity of ‘Greek’ and ‘Harry’ suggests a stronger effect of teaching on Harry in Example 6.13 (b) than (a), and the proximity of ‘the chair’ and ‘comfortable’ indicates the directness of the experience with the chair in Example 6.14 (b) but not in (a) (Lakoff and Johnson 1980: 130). Example 6.134 (a) I taught Greek to Harry. (b) I taught Harry Greek. Example 6.14 (a) I found that the chair was comfortable. (b) I found the chair comfortable. MORE CONTENT IS MORE (LARGER) FORM It is widely recognised that there is an iconic relation between the quantity of form and the quantity (strength, degree) of meaning. Namely, the more form, the more meaning (cf. Sapir 1921: 57–81).5 Both English and Japanese systematically use this grammatical metaphor to signal plurality, intensity and continuation. Example 6.15 (a) She tried and tried and tried and tried, but failed. (b) She tried, but failed. Example 6.16 (a) hanabana ﬂower.ﬂower

182 Metaphor and Iconicity

‘ﬂowers’ (b) hana ‘ﬂower’ In both Examples 6.15 and 6.16, (a) represents more amount of meaning than (b), navigated by the MORE CONTENT IS MORE (LARGER) FORM metaphor. In Japanese onomatopoeia, the [+Voiced] feature in phonology represents a bigger sound quality in meaning as in Example 6.17. Here again, the more form (i.e. voice added), the more meaning. Example 6.17 (a) tonton ‘knock knock’ (b) dondon ‘bang bang’ SYMMETRICAL MEANING IS SYMMETRICAL FORM The use of symmetrical form often implies a symmetrical relationship of the concepts represented. As linearity is an inescapable property of spoken language, we might speculate that a symmetrical relationship of concepts should be the most difﬁcult to represent iconically. However, most languages have some forms expressing symmetry, including simultaneous occurrence, alternative occurrence and mutual dependency of the concepts denoted. Symmetry is usually expressed in two dimensions: ‘ﬁrst, by distinguishing between co-ordination and backgrounding, and second, by overriding the temporal asymmetry of conjoined elements through the use of parallel diacritics of various types’ (Haiman 1985a: 102). Compare the pair of sentences in Example 6.18. Example 6.18 (a) The more he eats, the fatter he gets. (b) If he eats more, he will get fatter. Both (a) and (b) express roughly the same thing and any difference between the two is only subtle. However, (a) uses coordination to put the conjuncts in parallel, while (b) uses subordination to place the subordinate clause in the background. The two conjuncts in (a) are symmetrical whereas those in (b) are asymmetrical. Although there is temporal asymmetry of the two events in (a) as indicated by their linear iconicity, it is overridden by the use of parallel diacritics ‘the + com-

The Interplay of Metaphor and Iconicity in Grammar and Discourse 183

parative form of an adjective’. The two elements are lined up in parallel, so that they are taken to be equal in rank.6 ASYMMETRICAL MEANING IS ASYMMETRICAL FORM Asymmetrical relationships are generally more manifest in linear iconicity and proximity iconicity, in which either the sequential order or the distance of elements serve to denote the asymmetrical relationship among the content elements. Talmy (1978) observes another type of asymmetry in sentences that reﬂects a distinction between the cognitive–semantic categories of FIGURE and GROUND. Compare the pairs of sentences in Examples 6.19 and 6.20. Example 6.19 (a) The bike is near the house. (b) ?The house is near the bike.

(Talmy 1978: 628)

Example 6.20 (a) Tom resembles his father. (b) ?Tom’s father resembles John. It is clear that in neither case are the two sentences (a) and (b) synonymous. The asymmetrical relationship of the cognitive categories of ‘the bike’ and ‘the house’, and of ‘Tom’ and ‘Tom’s father’ makes (a) possible but not (b) in an ordinary context. The ‘bike’ and ‘Tom’ are the variable element or FIGURE; ‘the house’ and ‘John’s father’ are reference elements or GROUND. In simple sentences like the above, FIGURE is expressed as a topic and GROUND as a part of a comment. Hence, there is a diagrammatic correspondence between topic/comment representation and FIGURE /GROUND cognition. MARKED MEANING IS MARKED FORM Categorisation in language corresponds to the way in which we conceptualise the world. What cognitive linguistics claims about the relationship of linguistic categorisation and cognitive categorisation can also be interpreted in terms of diagrammatic iconicity. Lakoff (1987: 58) argues that ‘linguistic categories should be of the same type as other categories in our conceptual system. . . . Evidence about the nature of linguistic categories should contribute to a general understanding of cognitive categories in general.’ The basic premise of his methodology in utilising linguistic categories to understand cognitive categories presupposes that there is a correspondence between the two. This

184 Metaphor and Iconicity

correspondence can be taken as diagrammatic because the linguistic categories and the cognitive categories are ‘of the same type’. It is the sameness of type held between cognition and language that suggests structural diagrammatic iconicity.7 An important way in which the correlation between cognition and language is manifest is where a ‘basic/non-basic’ asymmetry in the cognitive categories is represented by a ‘marked/unmarked’ asymmetry in the linguistic categories.8 For example, a basic number, singularity, is expressed by no marker, whereas a non-basic number, plurality, is expressed by a marker in most languages. A basic tense, present tense, is semantically ‘unmarked’ and tends to have no overt marker, while non-basic tenses, past and future, are semantically ‘marked’ and tend to have overt markers. Thus, a cognitively basic or ‘unmarked’ notion tends to take an ‘unmarked’ form; a non-basic or ‘marked’ notion tends to take a ‘marked’ form.

TIME IS PLACE; CAUSE IS PLACE Metaphorical transfers that are more complex are observed crosslinguistically in vocabulary, in grammatical structures and in pragmatic discourses (cf. Claudi and Heine 1986, Sweetser 1990, Ohori 1991). Claudi and Heine (1986), in particular, have conducted an extensive investigation into the role played by metaphor in the evolution of grammatical morphology in Ewe. They claim, for example, that cognitive metaphors are useful to account for grammaticalisation in which a given lexical unit loses in semantic complexity and tends to be reduced to a grammatical marker, as illustrated in Ewe case marking. Metaphorical transfers abound in semantic change and grammaticalisation. One of the prominent grammatical metaphors comes from a SPATIAL–TEMPORAL–CAUSAL extension seen in polysemy and other grammatical constructions. The Japanese particle kara (‘from’), for example, displays a polysemy which can be explained by a metaphorical extension from PLACE to TIME, and to CAUSE. Example 6.21 (a) Watashi wa Tokyo kara ki-mashi-ta. [SOURCE place] I TOP Tokyo from come-POL-PAST ‘I came from Tokyo.’ (b) Kaigi wa san-ji kara hajimari-masu. [SOURCE time] meeting TOP three o’clock from begin-POL.PRES ‘The meeting begins at three.’

The Interplay of Metaphor and Iconicity in Grammar and Discourse 185

(c) Kore wa yoi hon da kara, yomi-nasai. [SOURCE cause] this TOP good book be from, read-POL.IMP ‘Because this is a good book, read it.’ In Example 6.21(a) and (b), kara is a postpositional particle, whereas in (c) the same item appears as a subordinate connective which conjoins two propositions. It is as if the subordinated proposition in (c) were taken as a spatial/temporal entity. SOURCE in a spatial domain is mapped onto a temporal domain in (b) and to a logical domain in (c). Consequently, the grammatical constructions produced by the polysemic metaphorical extension of the Japanese particle kara are motivated by the grammatical metaphors such as TIME IS PLACE, CAUSE IS PLACE and PROPOSITION IS PLACE. 6.1.3 Iconicity in discourse With prototypical examples in English and Japanese, the discussion above focused mainly on grammatical metaphors which relate to the correspondence between form and meaning in morphology, word formation, word order and syntax. This does not mean, however, that grammatical metaphors do not work in the correspondence between form and meaning in discourse. Rather, it means that there has been a great deal more literature dealing with the interplay of metaphor and iconicity in syntax than in pragmatics. One of the areas in pragmatics which looks promising for iconicity research is discourse analysis. In particular, various kinds of repetition displayed in conversation offer an ideal data source. Ishikawa (1991), for example, claims that conversational repetition serves as an iconic sign which links not only form and meaning, but also form and interactional function. Self-repetition and reduplication observed in conversation cue such iconic meanings as intensity, iteration and continuation. The interactional function that conversational repetitions display iconically is what she calls ‘a joint idea construction’, i.e. the identiﬁcation of the idea and stance towards that idea between participants. Tannen (1987, 1990) and Norrick (1987) examine repetition as a discourse-cohesive device in conversation. The types of repetition they have analysed in their data are repetition at similar intervals, sequence repetition, and a change from repetition to non-repetition. Although their main concern is functional rather than semantic, what they have discovered matches with the types of iconic meanings discussed in Chapter 2 (Section 2.4.3 in particular). Just as in poetic texts, repetition

186 Metaphor and Iconicity

in discourse can function as grammatical metaphors to guide or reinforce the textual interpretation. A non-exhaustive list of such grammatical metaphors of repetition includes: DOMINANCE or EMPHASIS IS QUANTITY; REGULARITY, UNITY or DEVELOPMENT IS ORDER; PAIRS MEAN COUPLING or CONTRAST; A FRAME or A CYCLE IS SAMENESS OF BEGINNING AND ENDING; and CHANGE IN CONTENT IS CHANGE OF FORM. Another important area in pragmatics for iconicity research is polite language. Languages of the world have developed various ways of expressing politeness. One of the most pervasive characteristics of such devices is that polite expressions tend to take a physically longer form than plain or neutral expressions. This is a special case of the grammatical metaphor, MORE MEANING IS MORE FORM, in which the meaning added by more form is politeness. Deference or politeness achieved by longer forms is felt as a pragmatic space between an interlocutor (or interlocutors) and a referent. Hence, DEFERENCE IS MORE FORM entails DEFERENCE IS DISTANCE. The following section will concentrate on this speciﬁc grammatical metaphor and discuss how it relates to polite expressions in Japanese and English.

6.2 Metaphor–icon links in polite language Shibatani (1990: 379) says in The Languages of Japan, ‘the honoriﬁc system appears to be ultimately explainable in terms of the notion of (psychological) distance’. This section elaborates, from the perspective of grammatical metaphors, the insight as expressed by Shibatani among many other grammarians (Brown and Gilman 1960, Tsujimura 1968, Yamanashi 1974, Brown and Levinson 1987, Nishida 1987, Kikuchi 1994, to list a few) into the relationship of deference and distance. It argues, through an analysis of Japanese honoriﬁc verb constructions with additional evidence from English, that the metaphor, DEFERENCE IS DISTANCE, motivates and underlies the structure and the use of the honoriﬁc system of language. DEFERENCE IS DISTANCE DISTANCE is essentially a spatial concept. In the basic understanding of space, there are two types of DISTANCE: HORIZONTAL DISTANCE and VERTICAL DISTANCE. It is widely observed that spatial and temporal cognition correlate with each other, and therefore one understands TIME as if it were SPACE, i.e. TEMPORAL DISTANCE. These three types of DISTANCE are expressed in linguistic terms in the choice of lexical items such as verbs, nouns, pronouns, adjectives, adverbs, prepositions, and in the gram-

The Interplay of Metaphor and Iconicity in Grammar and Discourse 187

matical tenses. What I am going to claim in this section is that expressions of spatial and temporal distance are also utilised as honoriﬁcs to represent a psychological or a pragmatic space perceived between an interlocutor and a referent, or between interlocutors. In other words, DEFERENCE is metaphorically conceptualised as VERTICAL, HORIZONTAL and TEMPORAL DISTANCE, as evidenced by metaphorical uses of linguistic expressions and grammatical constructions whose basic meanings involve spatial or temporal orientation. Languages differ as to what type of linguistic means they use to manifest DEFERENTIAL DISTANCE as well as what sort of DISTANCE they utilise, whether VERTICAL, HORIZONTAL or TEMPORAL. The following analysis will show that although both Japanese and English use all the three possibilities of DISTANCE, the most notable characteristic is that Japanese uses VERTICAL DISTANCE and English uses TEMPORAL DISTANCE.9 To illustrate, the data will be drawn from polite expressions relating to verbs in Japanese and English, because verb honoriﬁcation is syntactically the most complicated and the most productive. Japanese honoriﬁcation through verb forms is realised in two ways: by regularly formed grammatical constructions with the use of certain auxiliary verbs; and by lexical substitutes (suppletives) for neutral verbs. Regularly formed constructions apply to a fairly large number of verbs, while suppletives, substituting for the neutral verbs of daily use, form a closed class of about 50 verbs commonly used in modern Japanese. In English, verbs convey politeness through negation, interrogation, mood and tense. 6.2.1 VERTICAL DISTANCE Verticality of status assessment In sociocultural interaction, status is conceptualised by means of orientational metaphors. For example, a person of HIGHER STATUS is perceived as being in a HIGHER POSITION, and a person of LOWER STATUS in a LOWER POSITION, as evidenced in a number of English conventionalised metaphorical expressions (as underlined) as in the following example (Lakoff and Johnson 1980: 16). Example 6.22 (a) He has a lofty position. (b) She’ll rise to the top. (c) He’s at the peak of his career. (d) He’s climbing the ladder. (e) He has little upward mobility.

188 Metaphor and Iconicity

(f) He’s at the bottom of the social hierarchy. (g) She fell in status. Thus, HIGH STATUS IS UP; LOW STATUS IS DOWN. The Japanese language shares this basic metaphorical concept, as shown in the direct translation of the above examples (Watanabe et al. trans. 1986: 23–4). Example 6.23 (a) Kare wa takai chii ni aru. (b) Kanojo wa saikooi made nobori-tsumeru-da-roo. (c) Shoogai de ima ga zecchoo-ki-da. (d) Seikoo e no hashigo o nobot-te-iru. (e) Kare ni wa koojooryoku ga hotondo nai. (f) Kare wa shakai-kaisoo no teihen ni iru. (g) Kanojo wa chii o koroge-ochi -ta. It is assumed, therefore, that when the person addressed or referred to is socially superior to the speaker,10 this person is cognitively perceived to be at a HIGHER POSITION than the speaker. A higher position is produced by either the speaker’s raising the referent (or addressee) up, or by the speaker’s abasing his/her own position. Namely, RESPECT IS UP and HUMBLE IS DOWN. For example, in English we have: ‘His Royal Highness, the Prince of Wales’, in which ‘highness’, a word of vertical orientation, is used in the formulation of honoriﬁc titles to represent a VERTICAL DISTANCE between the referent, Prince Charles, and the speaker. Likewise, ‘to look up to’ and ‘to look down on’ are metaphorically used to mean ‘to respect’ and ‘to despise’. The grammaticalisation of verbs of vertical motion in Japanese In Japanese, VERTICAL DISTANCING is further manifested in the grammaticalisation of several honoriﬁc auxiliaries, such as itadaku (‘to receive from above’), sashiageru (‘to give upward’)11 and kudasaru (‘to give downward’).12 They have been grammaticalised ﬁrst by losing their original status as a full verb of vertical orientation and motion, and then by losing their status as a full verb (honoriﬁc suppletive) of donatory meaning, acquired by metaphorical transfer. This metaphorical transfer from vertical motion to donation13 is based on the conceptual metaphors, GIVING IS MOVING THINGS FROM ONE PERSON TO ANOTHER and RESPECT IS UP; HUMBLE IS DOWN. In modern Japanese, these three verbs, itadaku, sashiageru and kudasaru, are polysemic as they still maintain their original meanings of vertical orientation and

The Interplay of Metaphor and Iconicity in Grammar and Discourse 189

motion, and the honoriﬁc donatory meaning when they are used as honoriﬁc suppletives. In addition, when they are attached to the main verb as auxiliaries, they function as an honoriﬁc marker. Figure 6.3 represents this ﬂow of historical change in a vertical axis and the formulation of regularly formed honoriﬁc grammatical constructions in a horizontal axis. What is worth mentioning here is that the DEFERENCE IS VERTICAL DISTANCE metaphor not only manifests itself in semantic change and grammaticalisation, but also shows up in written language. The three verbs, itadaku, sashiageru and kudasaru, can be written either in hiragana

Verbs of vertical orientation and motion

I

itadaku ‘to put or have something on the top of something’

metaphorical transfer Honorific suppletives of donatory meaning

II

itadaku ‘to receive (HON)’

grammaticalisation

III

Verbs kaku + te ‘write + GER’

+

Honorific markers itadaku

‘HON (receive.from.above)’

Regularly formed constructions kai-te-itadaku ‘write-GERHON(receive.from.above)’

I

Fujisan wa itsumo choojoo ni yuki o itadai-te-iru.14 Mt. Fuji TOP always top at snow ACC have.on.the.top-GER-be ‘Mt. Fuji always has snow on the summit.’ (Neutral)

II

Mary wa sensei kara hon o itadai-ta. Mary TOP teacher from book ACC receive.from.above (HON)-PAST ‘Mary received (HON) a book from the teacher.’ (Deferential)

III

Watashi wa sensei ni suisenjoo o kai-te-itadai-ta. I TOP teacher DAT reference. letter ACC write-GER-HON (receive.from.above)-PAST ‘I received a favour from the teacher who wrote a reference letter for me.’ (Deferential)

Figure 6.3 DEFERENCE IS VERTICAL DISTANCE

190 Metaphor and Iconicity

(Japanese moraic syllabary) or in kanji (logographs). The logographs, (‘mountain top’), (‘up’) and (‘down’), used respectively in these three verbs, visually retain verticality in their iconic representation. A summary of the icon–metaphor links in the grammaticalisation of verbs of vertical motion in Japanese is presented schematically in Figure 6.4. The metaphor, DEFERENCE IS DISTANCE, works in the grammaticalisa-

Iconic Mapping

Metaphorical Mapping

Grammatical Metaphors DEFERENCE IS DISTANCE

Target

Source

kanji of vertical strokes

IMAGIC MAPPING vertical motion

deference

longer form

DIAGRAMMATIC MAPPING

FORM

MEANING

MORE DEFERENCE IS MORE FORM Grammatical Metaphors Figure 6.4 Metaphor–icon links in grammaticalisation of verbs of vertical motion

The Interplay of Metaphor and Iconicity in Grammar and Discourse 191

tion of honoriﬁc verbs in that vertical motion is mapped onto deference based on the generic image-schema of DISTANCE. There is an imagic mapping from the kanji used in these verbs of vertical motion to the input of source space. The blended space forges the metaphorical meaning of deference, which in turn leads to a diagrammatic mapping, navigated by the grammatical metaphor of MORE DEFERENCE IS MORE FORM. 6.2.2 TEMPORAL DISTANCE means that the use of past tense or future tense marks more deferential usage in some pragmatic situations. Temporal distance could be regarded as a special case of HORIZONTAL DISTANCE achieved by presenting an ‘irrealis’ encoding of the event. As mentioned earlier, the TIME/SPACE conceptualisation correlates, and thus renders TEMPORAL DISTANCE as a way to achieve DEFERENCE. This is a dominant strategy of distancing in polite formulaic expressions in English (Brown and Levinson 1987; Fleischman 1989). Examples 6.24–6.26 show that (a) is more polite than (b) because TEMPORAL DISTANCE is created by the use of future tense or past tense. TEMPORAL DISTANCING

Example 6.24 (a) Will you please come to the meeting? (b) *Do you please come to the meeting? Example 6.25 (a) Could we get together next weekend? (b) Can we get together next weekend? Example 6.26 (a) I was hoping we could get together next weekend. (b) I am hoping we can get together next weekend. Temporal distancing for honoriﬁc usage by past tense occurs very rarely in Japanese; nevertheless, it can occur in certain dialects in Hokkaido, Tohoku, Shikoku and Kyushu (Umegaki 1974: 185–6) as in Examples 6.27 and 6.28. Example 6.27 Toccha, e-da-ka-ne? Father, be-PAST-QUES-TAG ‘Your father was(HON) here, wasn’t he?’

192 Metaphor and Iconicity

Example 6.28 I-da, i-da Be-PAST, be-PAST. ‘He was(HON) here. He was here.’ The grammatical metaphor DEFERENCE IS DISTANCE not only works at the level of the metaphorical mapping of the syntactic and lexical domains onto the conceptual domain in honoriﬁcs. ‘Lengthening’ is indeed a factor to increase DISTANCE in the physical shape of linguistic expressions. All the forms of honoriﬁc usage discussed above have physically longer sequences than their non-honoriﬁc counterparts. This is a clear case of iconic manifestation in which MORE SPACE means DISTANCE. In this way, the grammatical metaphor DEFERENCE IS DISTANCE also holds at the level of diagrammatic iconicity.

6.3 Concluding remarks This chapter has argued that metaphor–icon links are manifest on all levels – phonological, morphological, lexical, syntactic and pragmatic – in synchronic and diachronic dimensions and in spoken and written modalities. The degree of iconicity may vary from example to example, from level to level, from language to language. It seems legitimate to assume that there is a continuum of bipolarity between iconicity and arbitrariness. Certain form–content relations are highly iconically motivated, others hold a highly arbitrary relationship, and still many others are in between these two poles (cf. Waugh 1992, 1994). There is also the problem of competing motivations. Those which compete with iconic motivations contribute to increasing arbitrariness. According to Haiman (1985a: 157–261), such competing motivations are largely economic in that they increase simplicity and manipulability of the linguistic (sub)system. Haiman exempliﬁes this tendency in the development of pidgins into creoles. There is an inverse correlation between the lexicon size and iconicity. The smaller the lexicon, the greater the iconicity. The greater the lexicon, the smaller the iconicity. For example, in the impoverished vocabulary, the relationship between antonyms is expressed explicitly by ‘x’ and ‘NEGATION+ x’ (e.g. gutpela and no + gut(pela) in New Guinea Pidgin). This is a case of diagrammatic iconicity signalling gutpela is the meaning shared because the form is shared. On the other hand, in the elaborated vocabulary, the same relationship is morphologically opaque as in ‘x’ and ‘y’ (‘agree’ and ‘refuse’) but more concise and manipulable (for a

The Interplay of Metaphor and Iconicity in Grammar and Discourse 193

further discussion of competing motivations, see Haiman 1985a, b, and Givón 1985). The discussion of iconicity and metaphor in general linguistic issues has predominantly focused on the spoken language in the literature explored in this section. Little attention has been paid to the written modality. This could be because major studies of iconicity have been done in alphabetic languages and not in logographic languages; and because iconicity has predominantly been treated as an issue in studies of syntax and not in studies of text. The next chapter will take up this neglected aspect of iconicity research with special reference to the logographic writing system of Japanese.

7 Iconicity and Written Language

Poetic thought works by suggestion, crowding maximum meaning into the single phrase pregnant, charged, and luminous from within. In Chinese character each word accumulated this sort of energy in itself. Earnest Fenollosa (in Pound 1936) This chapter focuses on the issue of written language from the following two viewpoints: the writing system and communication in graphic space. The case that I would like to make regarding iconicity and written language roughly corresponds with these viewpoints: the role of iconicity in the logographic writing system of Japanese; and the re-evaluation of written language in the context of visual graphic communication.

7.1 Written language and writing systems First of all, it can be pointed out that there is a general lack of serious discussion of iconicity in relation to written language. If we go back to the discussion of iconicity by Peirce, it is taken for granted that the concept of iconicity inevitably involves the visual dimension more predominantly than the auditory dimension. The examples that Peirce presented in his explanation of icons were visible signs such as a map, a ﬂoor plan, a portrait and a graph. Why, then, have we come to lose this focus when the discussion of iconicity has been put forward in the domain of language? Why did even the pioneers of iconicity research in the twentieth century, such as Jespersen, Jakobson and Bolinger,1 look for iconicity mostly in sounds and syntax and not in the written form? 194

Iconicity and Written Language 195

Even in the current recurrence of iconicity studies in cognitive and functional linguistics, the relationship between iconicity and written language has attracted very little attention. There seem to be two interrelated reasons. Firstly, spoken language in the modern Western context at least, is regarded as a primary phenomenon (e.g. Bloomﬁeld 1933: 21, Lyons 1972: 62, Fromkin and Rodman 1983: 3), and thus as a primary object of study. Secondly, the alphabet is taken as a paradigm example of the written language; and therefore, iconicity has come to be a secondary if not completely irrelevant issue in the discussion of writing systems. 7.1.1 The alphabet vs. non-alphabetic writing systems The primacy of speech and the alphabet Twentieth-century linguistics, inﬂuenced by structuralist thinking, treats speech as the primary and main phenomenon of language. Written language, on the other hand, is assigned a secondary status as a mere way of recording speech by visible symbols (cf. Bloomﬁeld 1933). This concept of the primacy of speech is closely related to the dominant writing system of major European languages, the alphabet. It is assumed that ideally there should be a one-to-one correspondence between alphabetic symbols and the sounds they represent. Although no existing alphabetic systems have yet ‘achieved’ this ‘ideal’ level, it is precisely in this tradition that modern linguistics has developed. Harris correctly argues that ‘modern scholarship has unquestionably and unquestioningly taken the alphabet as its central paradigm example of a writing system’ (1986: 37). He further insists that ‘the notion that in speaking we select the individual consonants and vowels which somehow emerge from our mouths threaded in the right order like beads on a string is simply the image of alphabetic orthography projected back on to speech production’ (1986: 41). Elsewhere (Harris 1989: 100), he suggests that the view of the alphabet as a more ‘advanced’ writing system than the non-alphabetic systems ﬁtted in with the doctrine of the primacy of speech, although this view has not gone unchallenged. Havelock (1976, quoted in Ong 1982: 28) even maintains that Greek analytic thought was a product of their introduction of vowels into the Semitic alphabet, which had consisted only of consonants and semi-vowels. For Havelock, the intellectual achievement of Greek civilisation was implemented by their analytic alphabet, which was devised to encode the sounds of their language.

196 Metaphor and Iconicity

Iconicity and non-alphabetic writing systems It is commonly believed in the Western tradition that the history of writing has progressed from pictograms to ideograms, ideograms to logographs, logographs to phonographs, phonographs to the alphabet. This European ethnocentric view of writing systems eventually degrades non-alphabetic writing systems, particularly non-phonographic systems, as being primitive. One cannot necessarily say that non-alphabetic writing systems are less advanced than the alphabetic system; but it is often in this context that iconicity is discussed in relation to writing systems, as a way to deﬁne the representation of pictographic and ideographic systems (Yule 1985: 162, Finegan 1994: 7–8, 480).2 In the pictographic system, the pictograms are mimetic to the objects they denote. Hence, pictographic signs for human beings, body parts, animals and everyday objects in the early Minoan period in Crete (cf. Diringer 1968), for example, resembled actual physical images of human beings, body parts, animals and other objects, as indicated in Figure 7.1. Ideograms, in theory, refer directly to notions. The Chinese character of refers to the notion of a tree when it is used as a radical. For example, , consisting of (‘tree’) and (‘to separate’), means a branch. However, it is not easy to distinguish ideograms strictly from pictograms, because some ideograms are pictographic, such as resembling the shape of a tree. In comparison to pictograms which manifest imagic iconicity, ideographic signs are generally more abstract and conventional, and thus display diagrammatic iconicity. A modern adaptation of pictographic and ideographic systems abounds in road signs, industrial design, and in electronic communications. For example, the telephone receiver sign resembles the actual object of a telephone receiver to be put vertically; and hence, it signiﬁes where the telephone set is located by way of metonymy. Likewise, the sign of a lit cigarette is mimetic to an actual image of a lit cigarette,

Figure 7.1 Cretan pictographic inscriptions in the early Minoan period (Diringer 1968: 57)

Iconicity and Written Language 197 :‘-(

Crying

:-)

Happy

:-*

Kiss

:-(

Sad

:-V

Shouting

,-)

Winking happy

Figure 7.2 Modern ideograms (James and Jansen 1995–2000: http://www. netlingo.com/smily.cfm)

and it stands for an area in which smoking is allowed. A moment’s reﬂection leads one to realise that even pictograms are highly abstract, conventional, context-dependent and culture-speciﬁc. But they are all iconic to the objects they depict. Over the past decade, electronic communication has spread out rapidly into the daily life of many people. Smileys (or emoticons) in Figure 7.2 are an illustration of modern ideograms mainly used in Internet and e-mail communications. The colon represents the eyes, the dash the nose, and the right or the left parenthesis the mouth. Hundreds of smileys have been invented and are widely used in order to convey positive and negative feelings. Smileys follow after the punctuation or in place of the punctuation at the end of a sentence, and indicate extra nuance attached to that sentence : –) Considering the fact that modern cultures use pictograms and ideograms extensively in linguistic and non-linguistic communication, it is nothing but a myth to regard pictograms and ideograms as ‘primitive’. They are iconic signs for writing, but not a primitive system of writing. 7.1.2 Chinese characters as ideograms and logographs The logographic nature of Chinese characters The most convincing evidence against the myth of the primitive nature of pictograms and ideograms is provided by the use of Chinese characters in the present-day writing systems of Chinese, Korean and Japanese. Earnest Fenollosa, as interpreted by Pound (1936), for example, acknowledges the motivated nature of Chinese characters, saying that ‘Chinese notation is something much more than arbitrary symbols. It is based upon a vivid shorthand picture of the operations of nature’

198 Metaphor and Iconicity

(Pound 1936: 8). Though he is aware of the fact that the pictographic root of many characters cannot now be traced, he thinks that the ideographic nature of Chinese characters is advantageous as a medium of poetry. Not only does he recognise the pitfall of the primacy of speech based on the premise of phonetic symbols, but he also sees the importance of written language, particularly ideographic language, in which the accumulated treasures of a cultural heritage are real and active with a visible etymology. Although there has been a controversy among sinologists over the pictographic nature of Chinese characters as to whether they are ideograms or phonographs (DeFrances 1984: 85), the strength of the phonographic argument could be interpreted as an illustration of the inﬂuence of the Western myth stated above. For example, an advocate of ‘phonograph theory’, DeFrances (1984: 129) admits that the inventories of both syllabic signs and radicals in the Chinese writing system are ‘defective’ in that they are not as well organised as the English alphabet. The vast majority of Chinese characters are logographic (i.e. corresponding to words), and less than 10 per cent are purely pictographic. These logographic characters often consist of a combination of radicals – phonetic constituents and the semantic constituents. While the phonetic constituents remind the reader as to how the characters are to be pronounced, the semantic constituents can be explained as ideograms because they refer to certain concepts. When the characters are decomposed into constituents, each constituent can have the potential to be explained graphically as it is often the case that they originated from pictographs or contain pictographic elements. Ma (1997) reports interesting psycholinguistic evidence whereby the visual elements in the Chinese writing system result in the frequent use of visual imagemappings in composing metaphors and naming objects when being compared to a counterpart context in English. Cognitive differentiation between logographs and phonographs In Chinese, only Chinese characters are used for writing. By contrast, the Japanese writing system combines both logographic (kanji) and phonographic (hiragana and katakana) systems (for details, see Chapter 4). Because of the mixed nature of the Japanese writing system, it offers further evidence in a clearer fashion in support of Fenollosa’s insight and Ma’s ﬁndings that logographic characters and phonographic notations are perceived differently, and that the cognitive process of logographs is motivated more by visual orientation.

Iconicity and Written Language 199

Psycholinguistic studies such as Haga (1971: 32–4) demonstrate that kanji and hiragana are cognitively different in that kanji are perceived as ‘large’, ‘heavy’, ‘strong’ and ‘tense’, whereas hiragana are ‘small’, ‘light’, ‘weak’, ‘lax’ and ‘beautiful’. Paradis et al. (1985), on the other hand, have surveyed the literature on the studies of dyslexia reported in clinical and experimental situations since the beginning of the twentieth century. They offer an account and an analysis of manifold cases of dissociation between logographic and phonographic scripts in reading and writing among Japanese dyslexia patients. Both psycholinguistic and neurolinguistic ﬁndings suggest that visually and cognitively Chinese characters are perceived and processed separately from phonographic characters. It can be concluded that the ‘phonograph theory’ of Chinese characters can be rejected on the basis of these cognitive studies. For the purpose of exploring further the ideographic and logographic nature of Chinese characters in the context of iconicity, the Japanese writing system provides a pivotal case because it combines both logographic and phonographic systems. In Japanese, kanji, Chinese characters similar to the ones presently used in Taiwan and Hong Kong, have remained as a medium to convey meaning in a logographic mode, while hiragana and katakana, two other syllabary systems invented by the modiﬁcation of some 50 kanji, are phonographic. In what follows I will discuss, with an illustration of the writing system of kanji, how an understanding of writing systems relates to the generation and interpretation of the texts. It will be clariﬁed how kanji as iconic signs can play a cognitive role both in the formation of the lexicon in Japanese and in the creation and interpretation of Japanese written texts.

7.2 Iconicity and kanji In the mixed writing system of Japanese, kanji function as graphic icons at both the generic and speciﬁc level. At the generic level, the overall graphic shape of kanji yields associations based on imagic iconicity. Because their graphic shape is more angular and distinctive than the shape of hiragana, they are associated with supposedly masculine characteristics such as ‘large’, ‘heavy’ and ‘strong’. Hiragana on the other hand have feminine associations such as ‘small’, ‘light’ and ‘weak’. Kanji also manifest diagrammatic iconicity on the generic level. Because of their semantic integrity, kanji are used to express content

200 Metaphor and Iconicity

words, which generally bear a greater semantic burden than function words. Hence, there is a diagrammatic correspondence between the visual complexity of kanji and their semantic density. In this way, kanji often constitute the ‘ﬁgure’ and hiragana the ‘ground’ on the page of a text. At the speciﬁc level, each kanji can be seen as a distinctive icon. The degree of iconicity varies from one character to another. Generally speaking, the relatively fewer, more basic and frequently used characters are pictograms, made up of simple and compound logographs. The vast majority of characters, less frequently used, are a compound form comprising an ideographic constituent and a phonographic constituent. Therefore, some characters retain a strong connection with their pictographic roots, while most other characters are only partially ideographic. In referring to kanji as logographs, we mean that the majority of characters correspond to words (and parts of words); and yet, the decomposed parts of kanji are pictograms, ideograms or logographs. Iconicity, therefore, plays an important role in the generation of their meaning. In the following analysis, the interplay of metaphor and iconicity is shown to be inﬂuential, to varying degrees, both in the formation of kanji and in word formation by kanji. 7.2.1 Metaphor–icon links in kanji Iconicity in the formation of kanji Kanji are classiﬁed into four categories based on the character formation process:3 (1) shookei-moji (‘pictograms’); (2) shiji-moji (‘simple logographs’); (3) kaii-moji (‘compound logographs’); and (4) keisei-moji (‘phonetic–ideographic characters’). Shookei-moji (‘pictograms’), the most basic scripts, maintain their close connection with original pictograms and, therefore, they are mimetic to the things they represent, as indicated in Figure 7.3. In these characters, there is an imagic iconic mapping between the shape of kanji and the concrete object (and the meaning) that they depict. Shiji-moji (‘simple logographs’) are the smallest in number and probably the closest to ideograms as they represent abstract ideas such as spatial orientation of (‘up’) and (‘down’), and basic numerals, (‘one’), (‘two’), (‘three’) and (‘ten’). However, they are also logographs as they directly correspond to the words expressing the concepts, as represented by the graphic shape of the characters. There is a

Iconicity and Written Language 201 Early Form

Modern Form

Meaning

‘gate’

‘the moon’ Figure 7.3

Shookei-moji of ‘gate’ and ‘the moon’

diagrammatic mapping between the form of these kanji and the concept that they represent. Kaii-moji (‘compound logographs’) consist of two or more constituents or radicals, which are pictograms or ideograms themselves and contribute to the meaning of the whole. For example, (‘bright’) consists of (‘the window’) and (‘the moon’); (‘old’) consists of (‘ten’) and (‘mouth’). Compound logographs represent their meaning by way of metaphor and iconicity. Sometimes, the pictographic root of radicals in compound logographs is traceable and one can easily reconstruct the etymology of the character. For example, (‘bright’) consists of (‘window’) and (‘the moon’). The meanings of the radicals constitute input spaces for the blend by means of iconicity and metonymy. As shown in Figure 7.4, the meaning of ‘brightness’ comes from the blend of the pictograms of ‘a window’ and ‘the moon’. The ‘moon’ then functions as a metonymy for ‘moonlight’. The blend mixes the meaning of ‘window’ and ‘moonlight’, and produces ‘moonlight through the window’. The blend further elaborates a more abstract meaning of ‘brightness’, based on the CAUSE–EFFECT metonymy. For the case of (‘old’) consisting of (‘ten’) and (‘mouth’), a similar but slightly more complicated explanation is involved. The meaning of comes from the blending of the two input spaces, represented by its radicals, the ideogram of ‘ten’ and the pictogram of ‘mouth’. The input of ‘ten’ functions as a metonymy for ‘ten generations’, and hence, it acquires an extended meaning ‘a long duration of time’. The input of ‘mouth’ is a metonymy for ‘speaking’. The blend mixes the meaning of ‘a long duration of time’ with another input meaning of ‘speaking’ and produces the meaning ‘something spoken over ten generations, or for a long duration of time’. Based on the idea

202 Metaphor and Iconicity

a window

the moon

opening

IMAGIC MAPPING

the moon

window IMAGIC MAPPING DIAGRAMMATIC MAPPING

moonlight

light brightness

moonlight through window

FORM

MEANING

EFFECT IS CAUSE Grammatical Metaphor/Metonymy Figure 7.4 Metaphor–icon links in

of ‘something spoken over ten generations’, the blend further creates a more abstract meaning of ‘old’. The fourth category of kanji, keisei-moji (‘phonetic–ideographic characters’), includes the majority of characters. They comprise two or more constituents, which are often pictograms or ideograms. One of the constituents stands for the sound of a whole character of which it is a part, while the other constituent(s) mainly signal meaning. In the same way as for compound logographs, all constituents including the phonetic ones usually contribute to the generation of meaning. For example, (‘to hear’) is expressed by the semantic constituent (‘ear’) and the phonetic constituent [moN], (‘gate’).4 Likewise, (‘to question, to inquire, to ask’), is expressed by the semantic constituent (‘mouth’) and the same phonetic constituent [moN], (‘gate’). In each case, both elements contribute to the generation of meaning. Representation of meaning in the ‘phonetic–ideographic characters’ can also be explained by way of metaphor and iconicity, as illustrated graphically for the meaning production of (Figure 7.5). As the pictographic root of radicals and is traceable, so the etymology of the

Iconicity and Written Language 203

gate IMAGIC MAPPING

ear listening hearing body part cavity

exit entrance

IMAGIC MAPPING DIAGRAMMATIC MAPPING

sound hearing Blend

FORM

MEANING

SPECIFICATION IS JUXTAPOSITION Grammatical Metaphors Figure 7.5 Metaphor–icon links in

character becomes visible. There is an imagic mapping from the visual form of kanji to each input mental space constituted by its radicals, a gate and an ear. The meaning of is activated in the blending of these two input spaces. Certain features of the inputs, such as ‘entrance’ in the ‘gate’ input space and ‘cavity’ and ‘hearing’ in the ‘ear’ input space, are blended into a larger structure of , in which a productive speciﬁcation or scenario is emergent – something like ‘hearing is “for sounds entering into the ear cavity” ’. The same analysis also applies to , in which the feature of ‘exit’ in the ‘gate’ input space is activated at the time of the blend to forge the meaning ‘questioning is “for sounds exiting out of the mouth cavity” ’. At the time of the blend there is also a diagrammatic mapping from the blend to the form, , assigning the meaning of ‘hearing, or receiving sounds in the ear’ or to the form, , ‘questioning, inquiring, or putting sounds out of the mouth’ by way of the grammatical metaphor of SPECIFICATION IS JUXTAPOSITION. Juxtaposition of ideographic and pictographic radicals is a key aspect of the formation of kanji. In general, there are three geometrical

204 Metaphor and Iconicity

patterns of juxtaposition: (1) left–right (e.g. ); (2) up–down (e.g. ); (3) enclosure (e.g. ). Whichever pattern a certain character takes, the meaning generation is manifested as a conceptual integration through creative blends of constituents. In the same way as Turner and Fauconnier (1995a) argue about the non-compositionality of formal expressions (such as ‘dolphin-safe’ and ‘child-safe’) in English, the meaning of kanji in Japanese is not compositional. There is no ﬁxed or standard way of reconstructing the meaning from the constituents alone. Every kanji tells a different story or scenario, which must be discovered imaginatively by the pictographic cues, inferences and background knowledge at the time of the blend. Iconicity in word formation by kanji Kanji are important not only as units of the Japanese writing system, but also as essential elements in word formation. Japanese is an agglutinative language in which new words are formed by putting morphemes together. These morphemes retain their original forms and meanings with little change during the combination process. Each kanji represents one morpheme, and functions as a highly productive element in the process of word formation. Similar to Latin, Greek and Anglo-Saxon roots that function in word formation in English, kanji are combined with other kanji to form numerous compound words. In contrast to English, however, in Japanese word formation by kanji, the semantic link is visible among words sharing the same kanji character. In addition, kanji are highly productive in that a few thousand kanji can be combined to produce hundreds of thousands of compound words. This means that an enormous number of compound words in the Japanese lexicon are diagrammatically linked by the grammatical metaphor SIMILAR MEANING IS SIMILAR FORM. Sharing the same kanji as a segment in the compounds signals that these compounds have similar shades of meaning. For example, the character for water, , is combined with other characters to produce the compounds in Example 7.1. Example 7.1 kanji pronunciation (a) (b) (c)

[kjo:suibjo:] [suizoQkaN] [mizudori]

semantic composition fear-water-illness water-family-building water-bird

English equivalent hydrophobia aquarium water bird

Iconicity and Written Language 205

Although the pronunciation of varies as [sui] in Examples 7.1(a) and (b) and [mizu] in 7.1(c), the semantic link of these three words is visibly and iconically obvious. Notice that in their English equivalents, ‘hydrophobia’, ‘aquarium’ and ‘water bird’, the semantic link is somewhat obscured because they are represented by three different forms, ‘hydro’, ‘aqua’ and ‘water’, derived from three different roots, Greek, Latin and Anglo-Saxon. The meaning of compound words written in kanji is often selfevident, because each kanji of a compound word conveys a distinct meaning and the combination results in semantic transparency as in the case of Examples 7.1(a) and (c). Sometimes, however, the combination of kanji does not directly convey the meaning. This is the case in Example 7.1(b), in which the meaning generation requires inferences and pragmatic knowledge about the concepts expressed by the characters. The blending model is useful in giving an account of this kind of meaning generation in compound words as well.5 Similar to the mechanism of meaning creation by kanji as a blend of radicals, a compound word can be treated as a blend of words represented in kanji. , for example, is a blend which forges its meaning by developing and exploiting the inputs of and , and then those of and , in addition to inferences derived from the inputs. (‘water’) functions as a metonymy for ﬁsh and animals living in water. (‘family’) extends its meaning by a metaphorical transfer to signify ‘a class of things with common features’, and it functions as a sufﬁx, when it is attached to . Hence, means ‘a class of animals and ﬁsh living in water’. Certain pragmatic knowledge comes in here, too, because for the blend of and (‘building’) to make sense, one has to be familiar with a concept of exhibiting or viewing of ﬁsh and sea animals for pleasure, rather than eating or hunting these creatures. The dominant link of this meaning creation with an iconic representation of kanji is with the pictographic character of . In other words, there is an imagic mapping from the visual form of the word to the meaning ‘aquarium’ by means of . Furthermore, a diagrammatic link with all the words having in the paradigmatic inventory of the lexicon shows how the iconicity of kanji is deeply rooted in the essence of language structure and use.6 Any ordinary Japanese text employs a mixture of phonographic and logographic systems. The more intellectual the content is, the greater the amount of kanji is used. Unlike the discursive and analytic nature of phonographic systems, kanji has an immediacy of presentational

206 Metaphor and Iconicity

form. They are more noticeable and striking on the page. It is a common technique in fast reading in Japanese that one skims kanji on the page. Pronunciation is a secondary matter in this kind of reading, as kanji matter as a bearer of meaning. Indeed, kanji are an imagic icon in that the shape directly suggests meaning; it is also a diagrammatic icon in that a complex meaning is expressed in a complex form.

7.2.2 Kanji as a poetic medium Not only do kanji play an important role in conveying an iconic meaning visually but they also provide a unique mode of representation among the world’s many writing systems. The essence of such an ideographic mode of representation is the juxtaposition of separate entities so as to evoke a new matrix or constellation of meaning. This is demonstrated both by the formation of kanji and by word-formation by kanji, and explained for the most part by the interplay of metaphor and iconicity. It is an inevitable consequence of this stance that kanji offer endless creativity and imagination in visual and verbal arts, particularly in calligraphy7 and poetry. Fenollosa (Pound 1936) highly praises the ideographic nature of Chinese characters and Chinese sentences as an embodiment of true poetry. He claims there is an essentially poetic nature to Chinese ideographic characters in that they combine what is seen in pictographic potential (i.e. iconicity in the terminology of the present study) with what is unseen in the metaphorical process of meaning production. According to Fenollosa, metaphor is a process in which material images are used to suggest immaterial relations. What he sees in Chinese characters is that they visibly maintain the ancient roots of such metaphors in themselves. Based on this recognition, he insists that ‘the Chinese written language has not only absorbed the poetic substance of nature and built with it a second work of metaphor, but has, through its very pictorial visibility, been able to retain its original creative poetry with far more vigor and vividness than any phonetic tongue’ (Pound 1936: 24). The poetic text seeks an optimal use of kanji both as an iconic and a metaphorical manifestation in the creation and the interpretation of its meaning. My analysis conducted throughout this study of the role of kanji in haiku texts (see Chapters 1, 3 and 4) can be seen as an attempt to substantiate Fenollosa’s insight with the help of modern cognitive poetics and semiotics. Orthographical revisions regarding the use of

Iconicity and Written Language 207

kanji, in particular, present an ideal way in which to demonstrate his perception. In sum, iconicity of kanji is key to the full understanding of the Japanese writing systems; at the same time, it further provides profound insights into the mode of representation in a general theory of written language in addition to a theory of poetic creativity. What Fenollosa (Pound 1936) observed in the ideographic nature of Chinese characters is no less meaningful to Imagist poets than to a cognitive linguist who sees the interplay of metaphor and iconicity in the formation of kanji and its role in poetic texts. Based on the above considerations concerning a close relationship of writing systems with an understanding of the meaning generation of the written text, the next section will move the focus onto more general issues of written language.

7.3 A re-evaluation of written language This section considers the status of written language as a broader linguistic issue for the purpose of providing an extended deﬁnition of written language in the context of communication in graphic space. This will be done in two steps. The ﬁrst attempt is to offer a comprehensive view of written language as a mode of communication in its own right. The second attempt is to extend the deﬁnition of written language to include various uses of written signs for communication in graphic space. By doing so, I would like to discuss how metaphor and iconicity relate to the issue of written language in the context of visual graphic communication. 7.3.1 The comprehensive view Roughly speaking, there have been three views about the status of written language in linguistic theory: the primacy of spoken language; the primacy of written language; and the coexistence of both. After brieﬂy explaining the ﬁrst two views, and rejecting the binary opposition of spoken vs. written language that presupposes the primacy of one over the other, the third synthetic and comprehensive view will be presented. The primacy of speech The ﬁrst view, and the presently reigning view, holds that spoken language is basic and primary, and that the written or printed text is a

208 Metaphor and Iconicity

secondary phenomenon. Bloomﬁeld’s famous dictum, ‘writing is not language, but merely a way of recording language by means of visible marks’ (1933: 21) has indeed dominated in the treatment of written language in modern linguistics. Major introductory textbooks of linguistics, for example, almost unanimously state that language is primarily a system of sounds or an oral behaviour (e.g. Bolinger 1975: 468, Fromkin and Rodman 1983: 3). Lyons (1972: 62–3) lists the following four points in which speech is prior to writing: 1. Phylogenetic priority. All linguistic communities have or have had speech. All linguistic communities do not necessarily have writing. 2. Ontogenetic priority. Speech comes before writing in the normal acquisition of a native language. 3. Functional priority. Speech has a wider range of communicative functions. 4. Structural priority. There is a correspondence between the basic units of spoken language with those of writing. Even though the ﬁrst two points are genetically true, they do not necessarily prove the claim for the primacy of speech over writing. Genetic value may not be relevant in considering the relationship of speech and writing in adult literate society. The latter two points are arguably misleading, because it is questionable as to whether speech has a wider range of communicative functions, nor is there a correspondence between speech units and written units. There are possibly two background reasons for this primacy of speech doctrine; but such supportive reasons can be rejected on the basis of ideological misconception. Firstly, as mentioned earlier in this chapter, the primacy of speech has been supported by European ethnocentrism in which a writing system is identiﬁed with the alphabet.8 Secondly, it is an effect of the modern structuralist counterattack against the tradition of comparative philology, in which the primacy of written language had been endorsed for a long time. It was in written texts that philologists investigated the roots of Indo-European languages. However, it is also a naive idea for modern linguists to claim that spoken language is the subject of analysis, because such language, in an absolute and natural form, exists only empirically. The analysis could only take place after the original spoken language had been recorded by some device, and in practice, after it had been documented in some notations (Haberland 1994: 5061).

Iconicity and Written Language 209

As regards iconicity, the primacy of speech doctrine, which is Saussurean in essence, inevitably sees language as a system of arbitrary symbols. There is little room left for a fruitful discussion of iconicity in this tradition, in spite of the fact that there was a recognisable amount of experimental psycholinguistic research on sound-symbolism (Brown et al. 1955, Brown and Nuttall 1959, Miron 1961, Weiss 1963, Tanz 1971, Ultan 1978). It is quite an exception that some pioneering research by Jakobson and Bolinger (see Chapter 2) tried to extend the scope of phonological iconicity beyond the conﬁnes of sound-symbolism and onomatopoeia.

The primacy of writing A modern advocate for the primacy of the written language doctrine is found in the movement of deconstruction. The method of deconstruction is essentially post-phenomenological and poststructuralist. Derrida argues that the Western philosophical tradition, from Plato to Heidegger and Saussure, has degraded writing compared with the immediacy and vivacity of speech or voice (Sarup 1993: 36). Derrida (1976) suggests that this rejection of writing is an illustration of phonocentricism reinforced by the alphabetic writing system. Furthermore, such an ethnocentric attitude is in effect related to logocentricism, the belief that the ﬁrst and the last thing is the Logos, the Divine Mind. Logocentricism entails that language is authorised from outside. By contrast, Derrida insists that there is nothing outside of the text. The text is read so closely that ‘the author’s conceptual distinctions on which the text relies are shown to fail on account of the inconsistent and paradoxical use made of these very concepts within the text as a whole’ (Sarup 1993: 34). The attention is on the text; but the aim is on the deconstruction of it. This Derridean emphasis on writing over speech has thus reversed the primacy of speech doctrine. And yet, it has gone too far in the other direction to the extent that it asserts that aspects of literary works ‘follow from purely visible features of texuality’ (Brogan 1994: 280). Besides, for Derrida, writing does not refer to an empirical exercise but an abstract structure inhabited by the trace of the text (Brogan 1994: 41). The primacy of writing has stayed on the abstract ideological plane, leaving little room for substantive discussion of written language in actual human communication. Consequently, the issue of iconicity is beyond the scope of their interest.

210 Metaphor and Iconicity

The oral–literate equation In anthropology and education, the study of written language has also gained attention through the growing interest in the issue of literacy and orality (Ong 1982, Goody 1987, Olson and Torrance 1991, Olson 1994, among others). These studies are concerned with the differences of thought processes or structure of consciousness between oral and literate societies. They analyse how writing and print restructure consciousness differently from ‘primary oral cultures’ (Ong 1982: 20–77). In spite of the fact that these anthropological and educational studies try to deﬁne the issue as the ‘oral–literate equation’ (e.g. Havelock 1991[1989]), their treatment can be viewed as a reappropriation of orality in the world of modern literacy. Moreover, what Ong (1982: 84) and Havelock (1991[1989]: 22–6) consider as writing is prototypically and predominantly alphabetic systems. Consequently, their discussion is biased by Eurocentric ideology, and their evaluation of orality in modern literate minds is consonant with the primacy of speech in modern linguistics. It is problematic to devalue written language in order to claim the primacy of speech, or to overevaluate it in order to re-establish the primacy of writing. Probably what is needed is a more integrated way of looking at spoken and written language, situating both in a whole spectrum of human communication from a broader semiotic perspective, such as the one developed by Harris (1986, 1995, 1996, 1998b). The synthetic view: autonomy and interrelationship Against the backdrop of such a research tradition, I would like to view written language as an autonomous mode of representation, coexisting with spoken language, as well as an integrated mode of representation, correlating with many other activities. It is only possible from the synthetic viewpoint that one can legitimately consider the ‘autonomy’ and the ‘interrelationship’ of the two physical media, one phonic the other graphic, neither of which has any particular logical or ontological priority. In literate societies, spoken language and written language exist side by side; and, therefore, both must be evaluated from the full context in which they are used and integrated. Namely, there is no priority of speech over writing nor writing over speech, as it is the situation in which an activity of speaking or writing occurs and is comprehended that decides the priority.

Iconicity and Written Language 211

The ‘autonomous’ view sees writing as a mode of communication in its own right. It does not even presuppose that speech must play a role in it. Kanji are uniquely useful to demonstrate the validity of this claim. Since they are logographs, they can convey meaning even when one does not know how to pronounce them at all. Likewise, words spelled in a combination of kanji can make sense, quite independently of how they are pronounced. An extreme case of this sort of written communication is that a Japanese person with no knowledge of spoken Chinese might be able to communicate some basic meanings with a Chinese person by way of written Chinese characters. Of course, the use of the characters must be fully and appropriately contextualised; but in theory neither phonology nor syntax is at all necessary. One possible claim against the autonomous treatment of written language might be a rejection of the deﬁnition itself by saying that it is just written signs, and not written language. Harris (1995, 1998b) meets this criticism with the premise that the equation of writing with ‘glottic’ writing is both arbitrary and too narrow. He deﬁnes ‘glottic writing’ as ‘forms of writing that are speciﬁcally designed to be integrated with speech communication’ (1998b: 116), and presents a justiﬁable claim that ‘if we make the mistake of assuming . . . that glottic writing is the paradigm case, and furthermore that its essential function is to represent different forms of speech, rather than to integrate speech with other activities, then we shall end up with an entirely distorted picture of Homo scribens’ (italics in the original) (1998b: 117). This wider view of written language will be further elaborated below as an extended deﬁnition. Another aspect of the synthetic view of written language is the interrelationship or ‘hybrid’ of written and spoken language. Poster (1990) argues this interrelationship as a dominant phenomenon of computermediated communication. He claims that ‘the writer encounters his or her words in a form that is evanescent, instantly transformable, in short, immaterial. . . . The writer thus confronts a representation that is similar in its spatial fragility and temporal simultaneity . . . to the spoken word’ (Poster 1990: 111). Sorensen (1999: 21), though admitting that ‘computer communication is fundamentally a written medium’, also maintains that it has obscured the distinction between speech and writing. According to her, there are three major traits that computer-mediated communication shares with speech – the evanescent, interactive and informal nature. Electronic texts are evanescent. Unlike written texts on paper, oral and electronic texts are not ﬁxed either in content or format. Without

212 Metaphor and Iconicity

being recorded in some material form, both texts disappear as soon as they occur (see also Bolter 1989). A networked computer environment makes electronic communication interactive as well. When two or more people are logged onto computer terminals at the same time, interactive ongoing communication can take place in the form of a ‘chat’ or a ‘conference’. They can respond to each other just as in face-to-face communication. It is generally observed that written language in electronic communication has the spontaneity and informality of speech. Computer-mediated communication holds little regard for surface errors such as misspellings and lack of punctuation. Smileys and emoticons are devised to signal emotions, feelings, emphasis, doubt, surprise, and so forth, while asterisks, question marks and exclamation marks are used to indicate emphasis, intonation and surprise.9 The synthetic view, which entails the treatment of written language both as independent of and interrelated with speech, leads the discussion towards an extended deﬁnition of written language. It is a key issue to understand the status of the written medium in the age of electronic communication. 7.3.2 An extended deﬁnition of written language An integrational deﬁnition Harris (1998b: 121–3) gives a lucid deﬁnition of writing from the integrational linguistic perspective. He situates writing in a broad spectrum of human communication by saying that ‘writing as a form of communication shares many features with drawing, painting and others which typically involve the marking of surfaces and the organization of spatial relations’ (1998b: 121). It is characteristic of Harris to see the issue of written language not just on the level of vision but more broadly on the level of space. This extended and integrated view encompasses both visual and tactile messages such as signed languages and Braille writing. Harris differentiates writing from gesture on the one hand, and from drawing on the other. The crucial factor to distinguish writing from gesture is that the written forms have a temporal duration, which allows for reprocessing. Writing is distinguished from drawing on the basis of the processing phase of communication as well. The sign is processed in a speciﬁc order or relationship in writing, but not in drawing. Writing is a mode of communication based on the use of spatial relations, which can be reprocessed over time.

Iconicity and Written Language 213

Consequently, the extended deﬁnition of writing entails an integration of two sets of dichotomies: glottic and non-glottic writing; and discursive and presentational representation. Glottic and non-glottic writing The broad deﬁnition of writing includes both glottic and non-glottic writing. Glottic writing is verbal writing in the sense most commonly used. It is identiﬁed with any forms of writing, ranging from alphabetic writing to logographic writing (cf. Harris 1995: 59). Non-glottic writing is all other written signs, which can be interpreted as meaningful signs in the context in which they are used. According to Harris (1995: 145), ‘the potential range of non-glottic writing is inﬁnite’. Furthermore, both glottic and non-glottic writing are equally valid forms of writing in graphic space. They may be used in combination, or may also be combined with other forms of graphic communication such as size, shape, colour, movement, and conﬁguration in general. For example, mathematical texts are an example of non-glottic writing. Written signs used in mathematics, such as =, +, -, π, \, >, are quite commonly used in glottic writing, both alphabetical and nonalphabetical. Glottic forms are also used in non-glottic writing. Glottic alphabetic symbols such as X, Y, Z are employed in non-glottic mathematical writing. Discursive and presentational representation The second trait of the extended deﬁnition of writing is that it integrates discursive and presentational representation. In her seminal work on symbolism, Langer (1957) introduces a modal distinction between the discursive and the presentational for the purpose of making a case against the dominance of Western logic that sees all articulate symbolism as discursive (1957: 88). She claims the existence of non-discursive or ‘presentational’ symbolism. By analysing various symbolic forms in culture such as metaphor, ritual, myth, music and other art forms, she maintains that presentational symbolism is a normal and prevalent vehicle of meaning, and that it must be recognised as such along with discursive symbolism. For Langer (1957: 79–102), the discursive mode represents a linear and successive conﬁguration of digital elements. She thinks that language is a primary example of the discursive mode of representation. Although her deﬁnition, ‘language is in a strict sense essentially discursive’

214 Metaphor and Iconicity

(1957: 96), reﬂects structuralist thinking and cannot be accepted as it is, one must also admit that some elements in language are discursive. As regards written language, glottic phonographic representation, both by the alphabet and by moraic hiragana and katakana, for example, is essentially discursive in nature. They constitute discrete units to be laid out in a linear fashion no matter which direction, right to left, left to right, or top to bottom, they are presented in. The narrowest deﬁnition of written language, which is a reﬂection of the primacy of speech (and alphabet) doctrine as seen above, probably ﬁts here. However, even a quick look at kanji will be enough to criticise the above deﬁnition as being too narrow. Despite the fact that kanji are laid out successively to form a sentence, and in this sense their conﬁguration is discursive, the composition of the character of kanji itself manifests a different mode of representation. As discussed earlier, most kanji have pictographic or ideographic constituents in themselves, which are not discursive. Thus, kanji display a blend of discursive and nondiscursive modes of representation. Smileys and emoticons also present a similar phenomenon. They are used in a linear discursive fashion in a sentence, written in glottic forms in a similar way to the writing of alphabetic words; but the make-up of smileys themselves is nondiscursive. In Langer’s terminology, it is a presentational mode in which all elements are ‘understood only through the meaning of the whole, through their relations within the total structure’ (1957: 97). Visual forms are essentially presentational, because the relations determining a visual structure are grasped in one act of vision. What is involved here is a simultaneous, integral presentation. Not all glottic forms are discursive. Kanji provide convincing evidence. Not all non-glottic forms are presentational, either. Acronyms used in computer communication offer a good example for the discursive representation of non-glottic writing. Most acronyms are formed by combining the initial letter of the words they stand for, such as ‘ASAP’ and ‘BTW’. Generally speaking, glottic forms are more discursive than presentational, whereas non-glottic forms are more presentational than discursive. Obviously, iconicity correlates more with presentational representation than with discursive representation; but the point is that it is one of the crucial motivations of written language, which is deﬁned here as a means of communication in graphic space. In passing, it is worth noting here that Langer’s classiﬁcation of discursive and presentational modes of representation is not only relevant

Iconicity and Written Language 215

to written language but also to a general discussion of a theory of language on a more abstract level. Elaborating her argument of presentational symbolism in relation to language, Langer (1957: 141) says that ‘metaphor is our most striking evidence of abstractive seeing, of the power of human minds to use presentational symbols’ (italics in the original). Here, she admits that even in the discursive mode of linguistic representation, the mechanism of meaning generation by metaphor involves presentational symbolism. Langer’s insight has two important bearings in my discussion of the interplay of iconicity and metaphor as a general linguistic issue. Firstly, the distinction between the two modes of representation provides a valid frame of reference for the analysis of glottic and non-glottic writing. Secondly, what Langer deﬁnes as presentational symbolism of metaphor, i.e. ‘abstract seeing’, is comparable to what I claim as an iconic moment in metaphor. In other words, metaphor inevitably entails an imagic and diagrammatic iconicity, which is, in Langer’s terminology, presentational symbolism.

The interplay of metaphor and iconicity in writing When one accepts the extended deﬁnition of written language, the issue of iconicity comes in as a relevant and important motivation with which writing is actually formatted, processed and interpreted. Written language used as a means of communication in graphic space involves the use of ‘glottic’ forms such as phonographic and logographic symbols and characters, and the use of ‘non-glottic’ forms of various symbols, pictograms, mathematical notation, musical notation, and so forth. In addition to these forms, spatial or relational factors are also important. To list a few, they concern the use of different kinds of typeface, spacing, indentation and layout. All of them have visual, graphic and conceptual effects on their discursive and presentational modes of representation both in composition and interpretation. Writing involves the interplay of metaphor and iconicity to a varying degree. Some units of writing manifest imagic and diagrammatic iconicity. Others manifest metaphors. Kanji are an optimal example of the mixture of all the manifestations of iconicity (see Section 7.2). Smileys and emoticons also evidence the interplay of metaphor and iconicity at work in their invention and use. In the formation of a smiley : –), for example, there is an imagic mapping between the visual schematic shape of facial elements (eyes, a nose and a mouth) and the shape of

216 Metaphor and Iconicity

punctuation marks (a colon, a dash and a right parenthesis). These three elements of a human face function as a metonymy for a smiling face. Namely, there is a metonymic mapping from facial elements to a facial expression, a smile. When a sentence is punctuated with : –), then, this icon for a smiling face functions as a kind of metaphor for feelings expressed by a smile. Thus, it can be interpreted in a variety of ways depending on the context, meaning something like ‘don’t take what I just wrote too seriously’, or ‘I’m sarcastic about it’. Even more strongly than the manifestation in the formation of the units of writing, metaphor–icon links manifest themselves in typographical representation and conﬁgurations of units. How a certain text is actually written and read visually on the page is a part of its meaning. Hence, iconicity concerns the choice of the size, the font, the style and the colour of scripts, their conﬁguration and the use of the entire graphic space in which the written text is contextualised and interpreted. Grammatical metaphors, such as IMPORTANT IS LARGE, MORE MEANING IS MORE FORM, DIFFERENCE (SIMILARITY) IN CONTENT IS DIFFERENCE (SIMILARITY) IN FORM, discussed in Chapters 2 and 6, navigate the iconic interpretation of non-glottic elements and structures of the written text. It is inevitable that iconicity is a dominant motivation in any piece of writing, whether handwritten, typewritten, printed, ﬁlmed, televised, or processed and transmitted electronically on the screen of computers. Examples abound from advertisements, posters, newspaper and journal headlines, to ordinary documents typed or word-processed. It is probably difﬁcult to ﬁnd written texts that do not use any diagrammatic structuring in such a way that large fonts signify importance, a different typeface signals difference in function and the same size means equivalence. However, it seems that iconic motivations and manifestations are so common that they are taken for granted, become automatic and slip from one’s attention. At the same time, because non-glottic features have been neglected in the study of written language, there has been little discussion from a linguistic point of view on the subject of formal elements and structures of communication in graphic space. Harris (1995) and Kress and van Leeuwen (1996) are rare exceptions, although their main concerns are not iconicity per se. These rare works not only predict the growing importance of visual media in the coming century, but also indicate the need for serious study of written language in wider and manifold manifestations.

Iconicity and Written Language 217

Iconicity in electronic communication With the popularisation of electronic visual media, accelerated by the widespread use of personal computers, the concept of graphic space has gone through a radical change. The computer treats both glottic and non-glottic written signs as equally valid forms of writing. In addition to the phonographic keys put on an equal footing with numerals and punctuation marks on the keyboard, the inventory of clip art icons in the hard disk transform writing into a truly mixed mode of representation.10 An enormous number of new notations such as smileys, emoticons and acronyms are being created and dispensed for use. Writing by computer drastically extends the possibility of representation on a mass scale. In the age of the typewriter, it was generally difﬁcult to change the size or the font of letters typed on the page. With computer processing, it is just a click of a mouse to change the size, the font, the style, the colour and the layout of the document on the page. One can even add sound and image to the page. Writing by computer fuses discursive and presentational representation with such ease and sophistication. It is no longer in the hands of a few professional specialists but of ordinary mass users that the freedom of integrated writing media resides. McLuhan’s famous thesis, ‘medium is message’ (1994 [1964]: 7), expresses a powerful insight in this regard. He anticipated the radical changes in the way people communicate in the electronic age, saying that ‘the “message” of any medium or technology is the change of scale or pace or pattern that it introduces into human affairs’ (1994 [1964]: 8). Electronic communication certainly changes the scale and pace of communication. It is on a mass scale, and without barriers. It is immediate and with high speed. Writing in electronic graphic space may be described as ‘a new pattern of human communication that has been introduced with the medium of the computer as a technological extension of mankind’ (Sorensen 1999: 47). In conclusion, I would like to propose that in the new phase of electronic technology, iconicity is a dominant ‘pattern’ of communication if one borrows McLuhan’s terminology. Moreover, this pattern is preferred because it is regarded as ‘human-friendly’. As I said in the ﬁrst page of this study, the word ‘icon’ has even become a common vocabulary item in Japanese with the advent of mass computer culture. What one generally sees on the computer screen today is neither mathematical formula nor enigmatic commands written in a linear array of symbols. The discursive representation of old

218 Metaphor and Iconicity

computer ‘language’ has been replaced by the presentational representation of icons, which are preferred because they are more ‘friendly’ to human users. It is a very interesting evolutionary consequence that highly advanced modern technology ﬁnally ﬁnds iconicity more ‘human’ than arbitrary symbolicity. From the cognitive point of view, this is a natural consequence. Iconic and metaphoric operations entail connecting similar things together and understanding abstract things in terms of bodily experiences. That is to say, the machine is developing human-like attributes as it becomes equipped with the basic operations of the mind: the interplay of metaphor and iconicity.

8 Conclusions

Iconicity plays a fundamental role in verbal and mental evolution, in transforming antiquated beliefs into verbal categories; it is most adequate to convey with precision mental contents of high complexity. Ivan Fónagy (1999) As discussed in the preceding chapters, iconicity and metaphor are indeed the motivating force not only in the creation and interpretation of meaning in spoken and written texts but also in the evolution of new communication media, which the twenty-ﬁrst century is to witness. I would like to conclude this study with a summary of my main claims, which generally support Fónagy’s insightful statement (1999: 29), and a presentation of a few further research issues.

8.1 Summary of main claims 8.1.1 Theoretical contributions There are three theoretical contributions made in the study. Firstly and most importantly, it has been demonstrated by prototypical illustrations in poetic and non-poetic discourse that metaphor and iconicity are important motivations governing the relationship of form and meaning in language. Secondly, it has been claimed that metaphor and iconicity should be treated as an entwined phenomenon rather than separate phenomena in linguistic theory. Thirdly, the status of written language has been re-evaluated in relation to iconicity. 219

220 Metaphor and Iconicity

Motivated nature of language The fundamental claim made throughout this study is that language is essentially motivated by the interplay of metaphor and iconicity. As main data for analysis I have used a number of Japanese poetic texts, haiku in particular, in comparison with English counterparts. Speciﬁc poetic texts have been chosen to provide optimal illustrations for both elicitation and explanation of metaphor–icon links, on the levels of a single text, parallel texts, visual and auditory manifestations. Even though it can be said that metaphor–icon links are more overt than covert in poetic discourse as opposed to the situation in everyday language, there are a variety of degrees and combinations with which metaphor and iconicity manifest themselves. Some texts, though smaller in number, display pure iconicity in which the visual or auditory form itself mirrors the meaning. It is often the case that such texts were composed with the conscious intention of being interpreted that way. Actually, most texts show iconicity more subtly through metaphors in that metaphors are mainly produced and apprehended in most texts, and it is such metaphors that navigate the diagrammatically iconic interpretations of the textual structure. The metaphorical reading reinforces certain iconic interpretations prevailing in the textual form such as syntactic structures, phonological patterns, the parallelism of certain lexical items, or the choice of graphic characters, letters and layouts in writing. Extending the discussion from poetic to non-poetic discourse, I have also argued for the existence of grammatical metaphors, which lead the diagrammatic mapping between form and meaning of language in general. Looked at this way, iconic manifestations are notable in word order, word formation, grammaticalisation, semantic change, polite language, written language, and communication in graphic space at large. Clariﬁcation of the interrelationship of metaphor and iconicity The major problem in the treatment of metaphor and iconicity in the past literature in semiotics and linguistics is that metaphor and iconicity have not been studied in a fully related manner within the same frame of reference. Although Peirce deﬁned metaphor as a subtype of iconic signs along with image and diagram, metaphor has been treated quite separately from image and diagram in the major literature. Imagic iconicity has been researched as issues relating to sound-symbolism in phonology, psycholinguistics and anthropology, while diagrammatic

Conclusions 221

iconicity has been a major focus in syntax. Metaphor has acquired a central place in semantics, cognitive and functional linguistics, in which iconicity has been either neglected completely or treated as a subordinate issue to metaphor. By contrast, this study emphasises the importance of treating both in the same framework because metaphor and iconicity are essentially entwined with each other. On the one hand, metaphor entails iconicity. Imagic and diagrammatic mapping of image-schematic structures resides in the creation and interpretation of meaning in metaphor. On the other hand, iconicity entails metaphor. Grammatical metaphors navigate the diagrammatic interpretation of linguistic forms. Likewise, a metaphorical reading of the text reveals the iconic structure in the text. In brief, two types of interplay between metaphor and iconicity have been discussed: iconic moments residing in metaphor; and metaphor giving iconic interpretation to form. Redeﬁnition of written language Manifestations of iconicity uncovered by the contrastive analysis of Japanese and English texts inevitably involve the important issue of written language and writing systems, which has been neglected in modern linguistics. In comparison to English, Japanese writing systems are not only more complex but also more iconic, because of the use of logographs. Writing and reading in Japanese, one is constantly exposed to three different scripts. This gives rise to a stronger orientation towards visual iconicity manifested in Japanese poetic and linguistic tradition, in contrast to auditory iconicity in English. Given the fact that the logographic writing system plays an important role in Japanese to convey meaning independent of and interrelated with speech, the narrow deﬁnition of written language as a way of recording speech is rejected. Writing must be deﬁned broadly as communication in graphic space. This deﬁnition includes both glottic and non-glottic writing and discursive and presentational modes of representation. The interplay between metaphor and iconicity becomes a key issue when one evaluates written language in such extended dimensions. Furthermore, writing by computer ultimately involves iconicity. A broader choice of typography and easier management of graphic organisation, in addition to the ongoing invention of new iconic signs in electronic communication, promote wider and more varied manifestations of imagic and diagrammatic iconicity, as illustrated and discussed in the previous chapter. Just as the use of kanji in combination with hiragana

222 Metaphor and Iconicity

and katakana results in a stronger tendency towards iconicity at the visual and presentational level in Japanese written language, the new technological development of electronic communication makes every user more conscious of the visual presentation of the iconic message. It is clear that the broader and integrated evaluation of written language is both valid and necessary in the age of electronic communication, which is in many ways iconic communication. 8.1.2 Methodological contributions Methodologically the study incorporates the cognitive model of blending with a Jakobsonian structural analysis in order to account for the complex mechanism of creativity by metaphor and iconicity in English and Japanese poetic texts. The model of blending or conceptual integration has been modiﬁed for the purpose of presenting in a more speciﬁc fashion the workings of metaphorical and iconic mappings. At the same time, this study puts a Jakobsonian analysis of the structure of texts in the new context of the blending model. This wedding of structural analysis and cognitive explanation beneﬁts both approaches. Elaboration of the blending model The major methodological contribution of this study is that I have added to the model of blending proposed by Turner and Fauconnier a few speciﬁcations for explicating the workings of metaphor–icon links. There are two reasons for using the Turner–Fauconnier model instead of the Lakoff–Johnson standard model for analysing the data. Firstly, the new model can provide a useful tool with which to understand ‘creative’ metaphors, and thus it meets the criticism addressed to the standard model that it mainly deals with conventional conceptual metaphors and not ‘fresh or creative metaphors’ (italics in the original) (Toolan 1996: 86). The Turner–Fauconnier approach is ﬂexible and comprehensive as they assume that meaning creation is an ongoing mental operation with an emergent structure in the blend of input mental spaces. The model can therefore incorporate structures not only from the expressed entities (i.e. input spaces) but also from the unexpressed (e.g. pragmatic contexts, background prior knowledge, cognitive models, inferences and emotions). This makes it a very realistic and effective instrument for literary analysis. Secondly, the model of blending can handle, within the same framework, the cognitive projection activated by a variety of formal expressions other than metaphors, such as the juxtaposition of linguistic elements, other conﬁgurational structures and the lexical devices for multiple meaning.

Conclusions 223

With regard to the dynamic relationship between form and meaning in the creation and interpretation of the text, I have tried to deﬁne, more explicitly by the use of graphic speciﬁcations (see Figure 2.6), metaphorical and iconic mappings, which occur at the time of blending. It has been shown throughout detailed analyses of poetic and non-poetic examples that this model can specify which parts of the metaphorical process, whether the input, generic or blended spaces, relate to the imagic and diagrammatic mapping of form and meaning. At the same time, the emergent structure in the blend is of particular importance because it clariﬁes the mechanism of creativity. Composing and understanding the text is a process of making sense out of the stream of images, concepts, knowledge and feelings, emerging together; or in short, it is a process of making sense over the entire array of many mental spaces. When applied to analyse the formation of kanji, word formation by kanji and written texts in general, the model has clariﬁed how imagic and diagrammatic mapping relates to communicating certain representational meanings. Applicability of the model for a wide range of examples, from the most foregrounded cases of poetic texts to an evolutionary process of communication in graphic space, has demonstrated that this approach provides a valuable perspective on aspects of cognition in a broader human context. Structural analysis and icon–metaphor links Detailed structural analysis of the form of a text is a necessary condition to examine how the interplay between metaphor and iconicity manifests itself. With reference to the Jakobsonian approach, by looking at certain abstract patterns in the distribution of linguistic elements, such as distinctive features, phonemes, syllables, morphemes, words, phrases and sentences, the internal structure of the text has been scrutinised. When two elements in the conﬁguration, such as rhymed words or syntactic parallelism, are equated because of their formal similarity, and when there is also a semantic or conceptual link between the two elements thus equated, then it can be said that they exhibit an iconic manifestation: similarity in form signals similarity in meaning. This type of formal analysis has been criticised on the basis of its treatment of the text as an autonomous structure, and therefore its exclusion of the role of a reader. However, the use of Jakobsonian methodology in the cognitive approach to the text has not only met the criticism but also reﬁned the cognitive poetic analysis, which emphasises the conceptual processing of both the writer and the reader.

224 Metaphor and Iconicity

8.2 Further research issues The present study, I believe, has shown the potential of metaphor and iconicity, or the blending model, to develop the theoretical understanding of other issues in linguistics, semiotics, literary criticism, communication and cognitive science. I would like to conﬁne this discussion, however, to the following two issues: the reformulation of the concept of language from a structural system to an ongoing act; and the ecology of communication media. 8.2.1 ‘Languaging’: language in the making The dynamic nature of language The form and meaning of signs can only be static entities on an abstract, theoretical level. There is a logical necessity to develop a discussion on that abstract level in order to clarify the basic structure of sign systems. Peircean semiotics and Jakobsonian poetics are both mainly engaged in this abstract level of investigation. However, this does not necessarily mean that such abstraction is the only way in which to understand the relationship of form and meaning. In fact, form and meaning are subject to communicative acts in actual contexts. As the contexts, whether they are personal, social or cultural, are neither constant nor ﬁxed, the human subject interrelating with such contexts is never constant nor ﬁxed. The relationship of form and meaning, therefore, is never constant either. The production and interpretation of meaning by form are constantly open, creative and in ﬂux. Communication, expression and thinking take place precisely in such ongoing dynamic processes. The model of blending has great potential for providing good insights into this ‘dynamic process’ of meaning creation. The model can explain how mental spaces are set up, structured and linked in accordance with grammatical, contextual and cultural knowledge. As a move in discourse unfolds, a network of mental spaces is in the making. Metaphor and iconicity are important motivations to the development of this dynamic network of meaning. As discussed in this study, there are diagrammatic mappings between the mental spaces at the time that they are being connected and integrated to forge new meanings. From language to ‘languaging’ The emphasis here concerns a shift in the concept of language from a pre-set system to a communicative act in the making. In Becker’s (1995:

Conclusions 225

414–15) terminology, it is a shift from language to ‘languaging’, a shift ‘from an idea of language as something accomplished, apart from this activity we have shared, to the idea of languaging, as an ongoing process’. Isami (1994 [1974], 1994 [1990]) also criticises the semiotic concept of language as a structural system of symbols, which is detached from the creative life processes of human beings. His claim is that language is an integral act of life. He insists that we are ‘living’ language with our feelings, inspirations, memories, judgements, responsibilities, plans and goals, rather than just ‘using’ or ‘manipulating’ linguistic symbols as a tool. In his words (Isami 1994 [1990]: 46), ‘the essence of language is a creative act of human beings’ (my translation). In addition to Becker and Isami, there are a few important works to be mentioned in this connection. One is ‘emergent grammar’ by Hopper (1998), who radically reverses the common view of language in which grammar comes before discourse. Instead, he insists that it is discourse in its ongoing process that shapes grammar. In this view, grammatical ‘forms are not ﬁxed templates but emerge out of face-to-face interaction in ways that reﬂect the individual speakers’ past experience of these forms, and their assessment of the present context, including especially their interlocutors, whose experiences and assessments may be quite different’ (1998: 156). What Hopper claims has an echo in ‘integrational linguistics’ advocated by Harris (1984, 1995, 1996, 1998a, b), Davis and Taylor (1990) and Toolan (1996). Integrational linguistics sees language ‘in relation to the relevant totality of human behaviour (Harris 1998a: 17). In other words, language is ‘contextually embedded and . . . this contexualization is always open to change’ (Toolan 1996: 13). Language is both a product and a mechanism of the process in which ‘human experiences are constantly structured and restructured by the need to make sense of present events in the light of past events and vice versa’ (Harris 1998a: 19). It is a challenging issue to be developed in the future to situate the model of blending more fully in this dynamic view of language. As revealed by the poetic analyses throughout the study, the claim of the interplay of metaphor and iconicity bears theoretical and methodological implications for clarifying the creative negotiation of form and meaning. 8.2.2 Ecology of communication media Consideration of iconicity in language leads me to propose a new concept of ‘ecology of communication media’. This can be deﬁned as a

226 Metaphor and Iconicity

study of how communication media live and evolve with reference to their environmental contexts. Just as ecologists predict the effects of human activities on ecosystems, linguists should think seriously about the effects of communication media on human interaction. There are two promising areas of iconicity studies in this regard: communication by computer and non-verbal communication. Communication by computer It seems to me that technological advancement in communication in the twenty-ﬁrst century points towards a certain direction in which ‘human’ factors are of particular importance. As discussed earlier, the evolution of communication by computer indicates a transition from a linear and discursive conﬁguration of arbitrary symbols towards a holistic and presentational representation of iconic signs. In other words, user-friendly computers require iconic interface. This tendency can be interpreted as an ecologically sanctioned trend, because it meets the challenge of mutual intelligibility in global communication media. Being an immediate sign, and connecting things of similarity, icons are easier to understand than symbols; and therefore, they have more potential for communication across geographical, linguistic, social and cultural boundaries. In order for computers to survive as human-friendly communication media, their effects on ecological aspects of human interaction will be a key issue for study. Non-verbal communication Another promising area of study is non-verbal communication, which involves a larger amount of iconic and presentational media than speech communication. Although it is pointed out that more than 65 per cent of information is conveyed non-verbally (cf. Birdwhistell 1970), non-verbal communication has received minor treatment in the study of language.1 From the ecological point of view, more serious attention has to be drawn to the mechanism of non-verbal communication, because an understanding of non-verbal communication is of utmost importance not only for communication among peers but also for communication across cultures. It is a promising area for universal communication, because it inevitably entails iconicity and metaphor. Important things are said in a loud voice. Strong emotions are expressed with distinct and speciﬁc motion. This kind of iconic functioning of non-verbal communication

Conclusions 227

deserves detailed study, because it will reveal the fundamental ecological nature of human communication. The interplay of metaphor and iconicity as a human issue In this study, the interplay of metaphor and iconicity has been presented as a concept both for the characterisation and the analysis of language structure and process. It characterises the meaning generation and interpretation in various linguistic phenomena: word formation, grammaticalisation, semantic change, poetic discourse, signed languages and writing systems, to list a few. It has been suggested that poetic texts and the logographic writing system, in particular, serve as an optimal illustration for validating the claim of the motivated nature of language. The interplay of metaphor and iconicity is a concept for analysis as well. As an analytical concept, it intersects different media, such as spoken, written, gestured, discursive and presentational representations. With the advancement of electronic communication, the very concept of communication is perpetually changing its appearance. It will be a task of linguistics to encompass such changes, identify the proper scope of the issues, and to examine the inﬂuences of such changes upon human interaction. The discussion developed in this study along with the interplay of metaphor and iconicity will have implications for this issue of the changing appearance of the communicative act in the future, not only among human beings but also in the human–machine interface. The interplay of metaphor and iconicity essentially concerns the relationship of form and meaning in action. It is a creative and dynamic process of the human mind. How human beings represent what they think and feel is an eternal question in both the arts and the cognitive sciences. Metaphor and iconicity are an indisputable part of this representational process.

Notes 1 Introduction 1. In this book, the term ‘icon’ is used in a narrower semiotic sense as deﬁned below than is usually employed in computer jargon in which it may refer to any visual symbols. 2. The computer examples mainly concern Macintosh screen display. 3. Incidentally, the wastebasket on the Macintosh screen is used not only for discarding the documents or applications, but is also used for ejecting ﬂoppy disks or compact disks from the drive. The latter function contradicts our conceptualisation of a wastebasket to the extent that computer users often ﬁnd it difﬁcult or unnatural to follow this ejecting procedure for the disk. 4. Metaphorical expressions are illustrated in italics, and metaphorical concepts in uppercase letters. 5. This does not imply that there is not also the reverse process of conceptualisation, i.e. understanding computers in terms of the human mind – COMPUTER IS A MIND. We say, for example, ‘this computer has a large memory’, ‘His computer has a mind of its own’ and ‘My computer is stubborn.’ 6. Wescott (1971: 418) observes that among the three major forms of language – speech, writing and sign language – iconic elements are most readily recognisable in the third. Landsberg (1980: 98) also maintains that ‘sign language and writing, at least in its evolutionary perspective, exhibit very extensive and clear iconic elements’. See also Deuchar (1990) and Taub (2001). 7. A mora is a unit of timing. Each mora takes approximately the same amount of time to pronounce. For further explanation, see Chapter 5. 8. For further explanations given in English about Japanese haiku, see Blyth (1952), Yasuda (1957), Henderson (1958) and Shirane (1998). 9. Word-for-word translation is given by the author and not in Matsuo (1996 [1694]). My word-for-word translation is based on Matsuo (1966 [1694]) and Matsuo (1996 [1694]). The in-text reference with different years of publication indicates that the year in square brackets is a source or an original work and the year in parentheses is an access volume according to which the citation is made. 10. Some rivers have human male names such as Bando-Taro (‘place-male name’) for Tone River. Furthermore, rivers are prototypically metaphorised as snakes in Japanese idioms, e.g. kawa ga dakoo-suru (‘A river snakes’), kawa ga hebi no yoo-ni magaru (‘A river curves like a snake’), etc. 11. For the frequency of vowels used in Oku no Hosomichi, see Table 5.7. 12. This is not to deny, of course, the importance of the Hallidayan tradition of viewing language as a social semiotic (Halliday 1978, 1994a, Hodge and Kress 1988, among others). In this tradition, signs are never arbitrary, and ‘motivation’ should be formulated in relation to the sign-maker and the context in which the sign is produced, 228

Notes 229 and not in isolation from the act of producing analogies and classiﬁcations. . . . All linguistic form is used in a mediated, non-arbitrary manner in the expression of meaning. (Kress and van Leeuwen 1996: 7) 13. For further discussion on Peircean deﬁnitions, see Chapter 2. 14. There are a few exceptions. Henle (1958: 177) seems to be the ﬁrst that borrowed the notion of iconicity from Peirce and claimed that ‘there is clearly an iconic element in metaphor’. Ricoeur (1979 [1978]: 147) takes up Henle’s insight and suggests that a semantic innovation by metaphor ‘is not only schematized but pictured’, implying Peircean ‘diagram’ and ‘image’. Danesi (1995) deals with the image content of metaphor as iconicity, and reports a psycholinguistic experimental study. Taub (2001) is a detailed survey of icon–metaphor links in American Sign Language. Janney (1999: 953) regards the issue of metaphor and iconicity as a new direction of research in pragmatics by claiming the need to ‘start paying more attention to questions related to how pragmatic stylistic and rhetorical choices function as ﬁgurative gestures in speech and writing’. 15. For a more detailed deﬁnition of ‘cognitive linguistics’, see Ungerer and Schmid (1996), Lee (2001) and Croft and Cruse (2004). 16. ‘Blending’ actually covers a wider range of linguistic and non-linguistic phenomena than metaphor. For further discussion, see Chapter 2 as well as Turner and Fauconnier (1995a, b), Turner (1996), Fauconnier and Turner (1998, 2002), among others. 17. Blasko and Merski (1999: 124) also state that ‘haiku is a particularly good candidate’ for studying the cognitive processes involved in creative writing because of its extreme brevity combined with great depth of meaning, the use of vivid imagery for both composition and appreciation, and its wide accessibility. 18. For discussion about poetic and literal (everyday) language, see Wellek and Warren (1963 [1949]: 22–8), Leech (1969: 5–6) and Toolan (1996: 24–97). For the claim about the poetic nature of everyday language, see Friedrich (1986), Gibbs (1994), Isami (1994) and Janney (1999).

2 Methodological Framework 1. Even in the days of structural linguistics, there were a few linguists who were well aware of the issue of linguistic iconicity. See Hockett (1958), for example. 2. Jakobson (1971 [1965]) also deals with another non-arbitrary property, i.e. indexicality of artiﬁce, or the factual or causal contiguity between form and meaning. However, he devotes most of his discussion to the issue of iconicity. 3. The term ‘cognitive poetics’ was ﬁrst used by Reuven Tsur (1983), whose definition, however, ‘speciﬁcally excludes the cognitive linguistic research in conceptual integration, blending, and metaphor’ (Freeman 2000: 253). Although Tsur’s work (1992a, b, 1996) on the relationship of prosody and cognitive processes marks an important contribution, I use the term ‘cognitive poetics’ in a broader sense as addressed by Freeman (2000) to

230 Notes

4. 5. 6. 7.

8.

9.

10.

11. 12.

incorporate the cognitive theory of metaphor and blending in the analysis of the structure and content of literary texts. Cognitive poetics is a growing ﬁeld of literary study which embraces various approaches and analyses from the perspective of the workings of the human mind. See Semino and Culpeper (2002), Stockwell (2002), Gavins and Steen (2003), among others. Volume numbers and paragraph numbers in Peirce (1931–58) are indicated in the square brackets. Deacon (1997) elaborates this point in his neurological explanation of the evolution of communication systems. Table 2.2 is reproduced with a slight alteration from Hiraga (1994: 7). Figure 2.1 is based on the graphic representation of blending provided in the website for blending (cf. http://markturner.org/blending.html). For a more detailed explanation, see Fauconnier and Turner (1998: 269–72, 2002: 45–50). Turner and Fauconnier use the term ‘mental space’ in contrast to the term ‘conceptual domain’, employed by Lakoff, Johnson and other cognitivists. Mental spaces are small conceptual arrays put together for local purposes of action and understanding, while conceptual domain is a vast structural array that could not be made active in thinking (cf. Turner 1996). The CONDUIT metaphor is also prevalent in Japanese. However, there are at least a few other competing metaphors for language in Japanese. They are FLUID, TOOL and FOOD. See Nomura (1996) on FLUID metaphor, and Hiraga (1995b) on TOOL and FOOD metaphors. The terminology that I suggest here represents the nature of the two types of diagrammaticality more clearly than Haiman. He deﬁnes ‘motivation’ as ‘ways in which the linguistic form is a diagram of conceptual structure, and homologous with it’ (1985b: 2) and ‘isomorphism’ as ‘the tendency to associate a single invariant meaning with each single invariant form’ (Haiman 1985b: 4). The present classiﬁcation also secures the term ‘motivation’ for a more general use in which it signiﬁes the ‘non-arbitrary’ relationship between form and meaning. Ohori (1987) offers a critique of Haiman’s classiﬁcation of iconicity. Also see Jakobson (1971 [1965]), Bolinger (1977), Haiman (1985a, b), Waugh (1992, 1994), Hiraga (1994), among others. Figure 2.6 is reproduced with a slight alteration from Hiraga (2003: 322).

3 Manifestation of Metaphor–Icon Links: Prototypical Examples 1. A number of critical studies have been devoted to philological and rhetorical aspects of this poem. According to Westerweel (1984: 56), the most important works include Tuve (1952), Summers (1954), Boultenhouse (1959), Walker (1962), Bennett (1963), Hastings (1963), Brown and Ingoldsby (1972) and Higgins (1977). Of course, all of these studies have taken up the close link between form and meaning of the poem; but none has a mention of this link as a case of iconicity. From the perspective of cognitive poetics, Stockwell (2002: 67–70) analyses the poem in detail using the

Notes 231

2.

3. 4.

5.

concept of ‘action chains’, but does not go very far in relating their formal patterns and meanings with iconicity of the poem. There are other repetitions in the poem. For example, ‘became most’, ‘let me’, ‘this day thy victories’ and ‘the ﬂight in me’ are repeated in each stanza. These repetitions reinforce the content expressed. Surprisingly, the poem appears in horizontal form in many modern editions (see Herbert 1994 [1633]: 35, for example). There is a different version of the poem in The Complete Poetical Works of Percy Bysshe Shelley (1943). The basic claim made in this book also applies to this version. The choice of this particular poetic text is my own; but the choice of the English Romantic period was suggested by Donald Freeman (1978a). He says, the poets of the English Romantic period were concerned perhaps more than those of any other epoch with the role of poetic language in a theory of poetry. The Romantic era thus is a highly appropriate locus for an inquiry into the relationship between poetic syntax and poetic structure. That relationship is one of iconicity: syntactic patterns in these poems correlate in particularly rich and interesting ways with the statement that each poem makes. (1978a: 654–5)

6. 7.

8. 9.

10. 11.

12.

For an analysis of iconicity in English Romantic poems, see also Freeman (1978b) and Ross (2000). The word ‘sister’ is used in the text as an attributive noun modifying another noun ‘ﬂower’; thus, it is counted as an adjective in this analysis. The term ‘allegory’ is used to denote an interpretation of a text as if there were an ‘other’ sense to which it referred. Often external (historical, religious and cultural) evidence is required to interpret allegorical personiﬁcation (i.e. personiﬁcation of natural entities and events that perform a brief allegorical action) in haiku texts. For further discussion of allegorical interpretation, see Wilkinson (1997). Word-for-word translation is given by the author and not in Sato’s translation. There is another complete translation of Oku no Hosomichi by Nobuyuki Yuasa (Matsuo 1966 [1694]). In this book, however, I use Sato’s translation (Matsuo 1996 [1694]), which is more faithful to the Japanese text. These cognitive metaphors moreover underlie the main theme of the travel sketch as a whole. Basho’s words, ‘learn about pines from pines, and about bamboo from bamboo’ (Hattori 1973 [1776]: 547, my translation) teach that if a haiku does not arise naturally from the object, the object and its observer become two, and the observer cannot realise the feeling of the object, since the self intervenes. Hence, what a poet must do is to be one with what Basho calls zoka, that is, the poet must be fused with nature. Then one comes to real enlightenment. For Basho, haiku is a realisation of such a unity. The poem’s reading as a description of natural events suggests another analysis of global blend. The use of ya, a kireji (‘cutting letter’), in yuku haru ya (‘passing spring’), seems to support the notion that the poem has two major input spaces, because the kireji functions to divide the text into two parts,

232 Notes

13. 14. 15. 16.

17. 18. 19.

20.

so that yuku haru ya and the rest of the poem are put in equation (cf. Henderson 1958: 189). It is often explained (ibid.) that the feeling evoked by the part accompanied by the kireji, ‘passing spring’ in this case, is illustrated by or compared to the part which follows the kireji, ‘crying birds and ﬁsh’. Therefore, the crying birds and ﬁsh are the source input space, while passing spring (and the feeling it evokes) is the target input space. Yuku is here also secondarily a metaphor of DEATH (namely, FAREWELL TO LIFE). For further discussion of kanji and blending, see Chapter 7. For an explanation, see note 18 of Chapter 3. Yamamoto (1984 [1951]: 22) deﬁnes three major traits of traditional haiku as ‘greeting’, ‘humour’ and ‘improvisation’. Haiku etymologically means ‘humorous phrase’. It is a developed form of the ﬁrst 17 mora-stanza of the linked poetry called renku, in which 36 stanzas are compiled in a row by a group of people at a poetry party. Haiku acquired the traits of ‘greeting’ and ‘improvisation’ from renku. I owe this observation to Mark Turner (personal communication). This interior/exterior dichotomy also applies to another reading of futami, ‘to see the lid’, to be interpreted as ‘to see the external Shrine of Ise’. The pun is termed kakekotoba (‘hanging words’) and constitutes one of the traditional rhetorical devices of Japanese poetry, particularly in short poetic forms such as haiku and tanka. Kooin (‘the light and shadow’) is a metaphor for time in a Japanese proverb which means ‘Time ﬂies’.

4 Vision as Meaning 1. Peer (1993) investigates the role of typographic foregrounding in Western poetry. 2. The term ‘logographic’ is used instead of ‘ideographic’, because such characters correspond to words (or linguistic units) rather than ideas. In fact, the characters often represent parts of words as well as a whole in the case of kanji. Strictly speaking, therefore, the term ‘logographic’ is slightly misleading as well. In the absence of a more appropriate technical term such as ‘morpho-graphic’, it continues to be used (cf. Crystal 1987: 200). 3. For a typology of writing systems, see Sampson (1985). 4. The use of the English alphabet is becoming very recognisable in presentday Japanese. However, as the use is still restricted to road signs, advertisements, brand names and index markers, this study does not deal with the alphabet as a set of phonographic units in Japanese writing. 5. There are also diacritics in English, albeit of lesser signiﬁcance than in Japanese. Diacritics in Japanese include two symbols, ° and ≤, placed above the kana characters to the right as in [pa] and [ba]. They indicate the need for voicing or the shift from [h] and [p] to [p]. This means that the use of diacritics in Japanese is more signiﬁcant than English. However, the use of diacritics pales into insigniﬁcance in comparison with the thousands of kanji, that are of more importance in the Japanese writing system. 6. For further details, see Ohno (1958).

Notes 233 7. For neuropsychological evidence, see Morton and Sasanuma (1984). 8. This last example reminds us of the American poet, E. E. Cummings, who experimented with the creative use of typographical and punctuation symbols in English. For further elaboration of the grammatical uniqueness of Cummings’s poetry, see Fairley (1975). Lennard (1991) is a survey of the poetic use of lunulae (the marks of parenthesis) from late ﬁfteenth century to the present, offering a detailed case studies of six major English poets who exploited them: Marvell, Swift, Coleridge, Byron, Browning and Eliot. 9. For details, see Brogan (1994: 43–4, 96–8, 224–5, 335–7). 10. There is no comparable tradition of pattern poetry in Japanese literature. 11. For blending and metaphor, see Chapter 2. 12. For the composition of kanji characters, see Sections 4.1.1 and 7.2.1. For the inﬂuence of Chinese characters (or ideograms) on Imagism, see Pound (1936). 13. When a string of words in English corresponds to one word in Japanese, dots are used instead of spaces to show word boundaries. 14. Visual iconicity and the inﬂuence of the Japanese writing system will be further elaborated in the following section, which attempts to analyse visual iconic manifestations in conventional poetry. 15. There are very few analyses of the subtle manifestations of visual iconicity in English poetry. However, works by Ross (1981, 1989, 1990) deal with subtle visual patterns in modern American poets such as Frost, Stevens and Williams as well as some European and Latin American poets such as in Paz and Neruda. 16. The notion of stanza normally presupposes end rhyme, whereas the section does not. 17. I have chosen the poems in Oku no Hosomichi (Matsuo 1996 [1694]) for three reasons. Firstly, Oku no Hosomichi, written as a travel sketch, consists of a main narrative body, 50 haiku poems by Basho and a few other poems by other authors. The totality of the 50 haiku poems is considered an integrated text in its own right, as they conform to general principles of composition and structural congruence. Secondly, the analysis of the haiku revision process will contribute to the current controversy (Muramatsu 1988) over the authenticity of the major copies of Oku no Hosomichi (for further discussion on the palaeographical evaluation of the copies, see Hiraga 1995a). To trace the revision process, I have used Sora’s draft (1689), Basho’s draft (1997 [1690]), Sora’s copy (1694), Basho’s correction (1694) of Sora’s copy, Soryo’s copy (1694) and the anthologies of Basho’s haiku compiled until the wide publication of the printed copy (1698) of Oku no Hosomichi (Matsuo 1957 [1694]) (for a comparison of various versions, see Ohtani 1962, Kaneko 1973 and Muramatsu 1988). Thirdly, it was demonstrated in my pilot study (Hiraga 1987a) that Basho consciously and unconsciously used iconicity in the process of phonological, semantic, syntactic and orthographical revisions. 18. For an integral analysis of the revision of a haiku, see Chapter 3. 19. Texts of various revisions in this chapter are based on Kaneko (1973). 20. Only few kanji manifest pure imagic iconicity such as (‘mountain’) resembling the shape of the actual object. However, most kanji used in Japanese writing convey meaning in a diagrammatic manner, because each character

234 Notes consists of a few radicals which represent meaning iconically. Kanji in Japanese are different from Chinese characters used in Chinese in that in Japanese they are primarily semantic units as opposed to phonographic units, hiragana and katakana. Chinese characters in Chinese have two functions – phonographic and logographic – at the same time (Brogan 1994: 112). 21. The situation might be comparable to the process in English poetry of associating certain etymological connections of the words in the texts; but the visual impact is certainly very different.

5 Sound as Meaning 1. Japanese spoken in the Tokyo area is regarded as modern standard Japanese for the purpose of this research. 2. As Ladefoged (1982: 220–4) admits, there has not yet been an agreed deﬁnition of a syllable. 3. [Q ] represents a glottal stop, and [N] represents a moraic nasal phoneme. Both [Q] and [N] constitute a single mora in Japanese. 4. For further explanation of morae and syllables in Japanese, see Vance (1987: 2–3, 56–76). 5. See Ladefoged (1982: 33–4) and Shibatani (1990: 159) for phonemic structures of English and Japanese. 6. For further explanation on Japanese onomatopoeia and sound-symbolism, see Asano (1978), Hamano (1986) and Kakehi et al. (1996). 7. Wellek and Warren (1963 [1949]: 158) and Shapiro and Beum (1965: 6) list this poem as a typical example of an onomatopoeic poem. 8. For a more detailed analysis, see Hiraga (1990). 9. Some words in this list (e.g. tintinnabulation, swinging) have onomatopoeic and sound-symbolic effects at the same time. It should also be pointed out that onomatopoeia and sound-symbolism are context-dependent. For example, ‘rhyming’(II-21) or ‘rolling’(IV-39) has a sound-symbolic effect in the context where the word is rhymed with ‘chiming’(II-21) or ‘tolling’ (IV-41) respectively. ‘Sinking’(III-30) and ‘swelling’(III-30) have a soundsymbolic effect in the context where these words are used in a syntactically parallel format prevalent in the whole text of the poem. 10. This fact is important in that the similarity of parallel syntactic patterns is reinforced by their similar phonological trait, i.e. being onomatopoeic. 11. The interpretation of the effects of sound-symbolic elements is based on claims made in the following research on sound-symbolism: Jespersen (1964 [1921]: 396–411, 1933 [1922]), Shapiro and Beum (1965: 9–16), Leech (1969: 96–9), Jakobson and Waugh (1979: 177–231) and Waugh and Newﬁeld (1995). 12. Classiﬁcations of distinctive features used in this research are based on Jakobson et al. (1951) and Jakobson and Waugh (1979). 13. I owe this observation to Linda Waugh (personal communication). 14. When a string of words in English corresponds to one word in Japanese, dots are used instead of spaces to show word boundaries, such as ‘seep.pierce’.

Notes 235 15. Semi-shigure (‘cicada-shower’) is a common Japanese metaphor for the sound of cicada in terms of drizzling rain. 16. See Hiraga (1986, 1987a) for a more detailed analysis of this poem, which includes discussion on orthographical revisions. Tables 5.5–5.8 and Figures 5.3 and 5.4 are reproduced from Hiraga (1987a). 17. I owe this observation to John Robert Ross (personal communication). 18. For the inﬂuence of haiku on Imagist poets such as Ezra Pound and William Carlos Williams, see Coffman (1951) and Hughes (1960), among others. 19. The wheelbarrow can move; but the mobility is felt much greater with the chickens than the wheelbarrow in these sentences. 20. For an analysis of these pairs, see Ross (1986: 193). 21. For a more detailed analysis, see Hiraga (1988, 2000b). Table 5.11 and Figures 5.8–5.10 are reproduced from Hiraga (2000b). 22. Kokin Waka-Shu (Ozawa 1971 [905]) was the ﬁrst major work of kana literature, compiled by Tsurayuki Kino and others. This anthology contains 1111 poems divided into 20 books arranged by topics such as seasons, love, travel, mourning, congratulations, and so on. It was the model of tanka composition for a thousand years until the eighteenth century. For a poetic theory of the time, see Tsurayuki Kino (1971 [905]). 23. The ﬁrst line, hisakata no, is a makura kotoba (‘pillow word’), a poetic epithet or attribute for a word, usually occupying a ﬁve-mora line and modifying the ﬁrst word of the next line. Each pillow word, of which there are possibly a few dozen, is associated with certain concepts. The word hisakata, meaning ‘distance’, is always to be associated with light – it must be linked either with hikari (‘light’) in the second line, or less syntactically likely, with hi (‘day’) in the third line. 24. This feels ‘hidden’, because Japanese speakers naturally hear in terms of mora (corresponding to writing systems) rather than decomposed individual sounds. 25. Kawamoto (1991) has been a rare exception. I owe this observation to Haruo Shirane (personal communication).

6 The Interplay of Metaphor and Iconicity in Grammar and Discourse 1. ‘And’ itself does not have a temporal meaning. It is the order of segments combined by a narrative usage of ‘and’ that is iconic to the temporal ordering of the events being narrated. 2. In English, however, with an additional use of an adverbial phrase or a subordinate clause, the same content can be expressed in a reversed order, as in Example 6.10. Example 6.10 (c) John sat down after he came in. (d) Before John sat down, he came in. This is an issue of competing motivation where given two options of form, one can either use iconic motivation or other motivations such as economic or pragmatic.

236 Notes 3. The more nominal character the attribute has, the closer it stands to the head noun in both postnominal and prenominal modiﬁers (cf. Posner 1986: 315) as in: (a) a heavy English mahogany writing table (b) a heavy table for writing made of mahogany from England 4. Langacker (1987: 39) also points out that ‘grammar embodies conventional imagery’. What he meant by ‘imagery’ is equivalent to ‘metaphor’ in my usage, as he further elaborates it by saying that ‘[imagery] structures a scene in a particular way for purposes of linguistic expression, emphasising certain facets of it at the expense of others, viewing it from a certain perspective, or construing it in terms of a certain metaphor’ (ibid.). This image-metaphor is quite useful when explaining the subtle semantic differences between the following two sentences: (a) He sent a letter to Susan. (b) He sent Susan a letter. Langacker explains that because of the use of the preposition ‘to’ in (a), it highlights the PATH travelled by the letter with Susan as a GOAL, whereas (b) highlights the proximity of Susan and the letter, namely, the possession of the letter by Susan with the juxtaposition of the two nominals. 5. Counter-examples abound as well. The extreme case is the problem of zeroform, i.e. ellipsis and silence, which bear meaning. The issue, however, seems too vast and complicated to be properly handled here. 6. Still one cannot fully escape from the linear iconicity in Example 6.18(a) in which what is expressed by the ﬁrst conjunct occurs ﬁrst, and the second conjunct second. Compare (a) and (c) in that example. Example 6.18 (a) The more he eats, the fatter he gets. (c) The fatter he gets, the more he eats. The semantic difference between (a) and (c) depends on the linear iconicity of the two conjuncts. 7. Lakoff (1987: 113–14) lists four cognitive models by which human categories are characterised: propositional, image-schematic, metaphoric and metonymic models. It is assumed that the linguistic categories give evidence for the existence of these four models which, at the same time, reﬂect the nature of the cognitive categories. 8. Markedness is used in two ways in the literature: formal markers, e.g. overt markers vs. no overt markers; and markedness in semantics, pragmatics, cognition, etc. The second type of markedness, which I refer to alternatively as ‘basic/non-basic asymmetry’, is not necessarily represented by the ﬁrst type. It is also indicated by word order, contextual differentiation, and so on. 9. For a more detailed and comprehensive discussion about the metaphorical basis of honoriﬁc verb constructions, particularly with respect to HORIZONTAL DISTANCE, see Hiraga (1999c).

Notes 237 10. ‘A person socially superior to a speaker’ (Harada 1976: 503) here designates a prototypical person in an idealised cognitive model. In actuality, the deﬁning qualities of ‘socially superior person’ to a speaker may vary from culture to culture. For example in Japanese culture, not only social status but also seniority, gender, in-group/out-group orientation and the formality of the situation are the important factors in coming to a decision as to who belongs to the ‘socially superior’ category. Although these sociolinguistic factors are an important issue in the honoriﬁcs research, they are not an immediate concern of this study (see Ide (1982) for the sociolinguistics of the Japanese honoriﬁc system). 11. Sashiageru is derived from ageru (‘to move something from a lower position to an upper position’, ‘to give’) [8c.]. Ageru used to have both donatory and honoriﬁc senses, but it gradually lost its honoriﬁc meaning. Then it underwent a syntactic change by acquiring a preﬁx sashi-, whose function is emphatic of the verb meaning in both donatory and honoriﬁc senses. Therefore, in modern Japanese, sashiageru is a donatory verb with an honoriﬁc meaning, whereas ageru is a donatory verb without an honoriﬁc meaning. Time reference in brackets such as [8c.] indicates eighth century in which the ﬁrst documentation of the usage was made, according to Nihon Kokugo Dai-Jiten (1976). 12. Kudasaru is derived from kudasu (‘to move something from an upper position to a lower position’) [8c.]. Kudasu underwent a syntactic change, where it was passivised to kudas-are-ru, and then the sufﬁxes -are and -ru merged to make kudasaru [14c.]. 13. For details on giving and receiving verbs, see Kuno (1973: 127–35). 14. Verbs in question are boldfaced in the examples.

7 Iconicity and Written Language 1. Bolinger (1946) showed a genuine interest in the problems posed by the use of written language; however, he did not elaborate them within the framework of linguistic iconicity. Wescott (1971) is a rare exception. He argues that language contains iconic elements in its three major forms – speech, writing and sign language – claiming that iconicity is more readily noticeable in sign language than in writing or speech. 2. Although most graphonomists assert that alphabetic signs evolved from iconic signs or pictures (cf. Gelb 1963, Diringer 1968, among others), modern linguistics does not treat alphabetic signs as iconic. For a discussion about the import of the iconic elements of the alphabet, see Landsberg (1979). 3. There are additional two categories, tenchuu-moji (‘derivative characters’) and kasha-moji (‘phonetic loans’), based on usage. But they are rather few in number. 4. Kanji usually have more than one pronunciation. has [buN], [moN] and [ki], while has [moN], [to] and [toN]. 5. Shibasaki (2001) offers a blending analysis of compound words in kanji in Japanese and Chinese. 6. For an analysis of diagrammatic iconicity in English compounds, see Ungerer (1999).

238 Notes 7. Although calligraphy is outside the scope of this study, it is worth noting that there is a close relation between the development of calligraphy and the development of poetry in the Chinese and Japanese literary tradition. 8. McLuhan (1994 [1964]: 81–8) also points out the inﬂuence of alphabetic writing upon Western thinking. 9. Notice that the newly adopted devices of writing in computer communication – various combinations of punctuation marks and exclamation marks as ideographs – are oriented by the principle of iconicity rather than arbitrariness. 10. See Wyss (1999) for a typology of iconic manifestations in Internet communication.

8 Conclusions 1. McNeill (1992) pioneers the research of gestures, language and cognition, incorporating metaphor and iconicity. See also Armstrong et al. (1995: 22) for a discussion of iconicity as a precondition for ‘all kinds of communicative behaviour, both linguistic and nonlinguistic’.

References Akabane, Manabu (1970) Basho Haikai no Seishin [Spirit of Basho’s Haikai]. Tokyo: Shimizu Koobun-do. Akimoto, Fujio (1970) ‘Basho no Suikoo’ [Revisions of Basho]. In Hassoo to Hyoogen [Conception and Expression], ed. Genyoshi Kadokawa, 244–68. Tokyo: Kadokawa Shoten. Anderson, Douglas (1984) ‘Peirce on Metaphor’. Transactions of the Charles S. Peirce Society: a Quarterly Journal in American Philosophy, 20(4): 453–68. Arima, Michiko (1996) ‘Japanese Haiku vs. English Haiku vs. Concrete Poetry’. Poetica, 46: 137–52. Aristotle (1954) Rhetoric and Poetics. Trans. Rhys Roberts and Ingram Bywater. New York: Random House. Armstrong, David F., William C. Stokoe and Sherman E. Wilcox (1995) Gesture and the Nature of Language. Cambridge: Cambridge University Press. Asano, Tsuruko (ed.) (1978) Giongo Gitaigo Jiten [Dictionary of Onomatopoeic and Mimetic Expressions]. Tokyo: Kadokawa Shoten. Aso, Isoji (1961) Oku no Hosomichi Koodoku [Lectures on Oku no Hosomichi]. Tokyo: Meiji Shoin. Attridge, Derek (1996 [1987]) ‘Closing Statement: Linguistics and Poetics in Retrospect’. In The Stylistics Reader: From Roman Jakobson to the Present, ed. Jean Jacques Weber, 36–53. London: Arnold. Ayer, Alfred J. (1968) The Origin of Pragmatism: Studies in the Philosophy of Charles Sanders Peirce and William James. London: Macmillan – now Palgrave Macmillan. Beardsley, Monroe (1958) Aesthetics: Problems in the Philosophy of Criticism. New York: Harcourt, Brace and World. Becker, Alton. L. (1995) ‘Language in Particular: a Lecture’. In Beyond Translation: Essays toward a Modern Philology, 405–26. Ann Arbor: The University of Michigan Press. Bennett, Joan (1963) Five Metaphysical Poets, 3rd edn. Cambridge: Cambridge University Press. Birdwhistell, Ray L. (1970) Kinesics and Context: Essays on Body Motion Communication. Philadelphia: University of Pennsylvania Press. Black, Max (1962) Models and Metaphors. Ithaca, NY: Cornell University Press. Blake, William (1978 [1810]) ‘Vision of the Judgement’. In William Blake’s Writings, Vol. II, ed. G. E. Bentley, Jr, 1007–28. Oxford: The Clarendon Press. Blasko, Dawn G. and Dennis W. Merski (1999) ‘Haiku: When Goodness Entails Symbolism’. Metaphor and Symbol, 14(2): 123–38. Blyth, Reginald H. (1952) Haiku. Tokyo: Hokuseido. Bloomﬁeld, Leonard (1933) Language. New York: Holt, Rinehart and Winston. Bolinger, Dwight L. (1946) ‘Visual Morphemes’. Language, 22: 333–40. —— (1965) Forms of English. Cambridge, Mass.: Harvard University Press. —— (1975) Aspects of Language, 2nd edn. New York: Harcourt Brace Jovanovich. —— (1977) Meaning and Form. London: Longman. 239

240 References Bolter, Jay D. (1989) ‘Beyond Word Processing: the Computer as a New Writing Space’. Language and Communication, 9(2/3): 129–42. Bouissac, Paul, Michael Herzfeld and Roland Posner (eds) (1986) Iconicity: Essays on the Nature of Culture: Festschrift for Thomas A. Sebeok on His 65th Birthday. Tübingen: Stauffenburg Verlag. Boultenhouse, Charles (1959) ‘Poems in the Shapes of Things’. Art News Annual, 65–83. Brogan, Terry V. F. (ed.) (1994) The New Princeton Handbook of Poetic Terms. Princeton, NJ: Princeton University Press. Brooks, Cleanth and Robert Penn Warren (1960) Understanding Poetry. New York: Holt, Rinehart and Winston. Brown, C. C. and W. P. Ingoldsby (1972) ‘George Herbert’s “Easter Wings” ’. Huntington Library Quarterly, 35: 131–42. Brown, Penelope and Stephen Levinson (1987) Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Brown, Roger W., Abraham H. Black and Arnold E. Horowitz (1955) ‘Phonetic Symbolism in Natural Languages’. Journal of Abnormal and Social Psychology, 50: 388–93. Brown, Roger and Albert Gilman (1960) ‘The Pronouns of Power and Solidarity’. In Style in Language, ed. Thomas A. Sebeok, 253–76. Cambridge, Mass.: MIT Press. Brown, Roger and Ronald Nattall (1959) ‘Method of Phonetic Symbolism Experiments’. Journal of Abnormal and Social Psychology, 59: 441–5. Bybee, Joan L. (1985) ‘Diagrammatic Iconicity in Stem-Inﬂection Relations’. In Iconicity in Syntax, ed. John Haiman, 11–47. Amsterdam: John Benjamins. Claudi, Ulrike and Bernd Heine (1986) ‘On the Metaphorical Base of Grammar’. Studies in Language, 10(2): 297–335. Coffman Jr., Stanley K. (1951) Imagism: a Chapter for the History of Modern Poetry. Oklahoma: University of Oklahoma Press. Cooper, William E. and John Robert Ross (1975) ‘World Order’. In Functionalism, eds Robin E. Grossman, L. James San and Timothy J. Vance, 63–111. Chicago: Chicago Linguistic Society. Croft, William and D. Alan Cruse (2004) Cognitive Linguistics. Cambridge: Cambridge University Press. Crystal, David (1987) The Cambridge Encyclopaedia of Language. Cambridge: Cambridge University Press. Cummings, E. E. (1972 [1958]) ‘l(a’. In Complete Poems 1913–1962, 673. New York: Harcourt, Brace Jovanovich. Danesi, Marcel (1995) ‘The Iconicity of Metaphor’. In Syntactic Iconicity and Linguistic Freezes: the Human Dimension, ed. Marge E. Landsberg, 265–83. Berlin: Mouton de Gruyter. Deacon, Terrence W. (1997) The Symbolic Species: the Co-Evolution of Language and the Brain. New York: W. W. Norton. Deane, Paul (1995) ‘Metaphor of Center and Periphery in Yeats’ “The Second Coming” ’. Journal of Pragmatics, 24(6): 627–42. DeFrances, John (1984) The Chinese Language, Fact and Fantasy. Honolulu: University of Hawaii Press. Derrida, Jacques (1976) On Grammatology. Trans. Gayatri Chakravorty Spivak. Baltimore and London: The Johns Hopkins University Press. Deuchar, Margaret (1990) ‘Are the Signs of Language Arbitrary?’ In Images and

References 241 Understanding, eds Horace Barlow, Colin Blakemore and Miranda WestonSmith, 168–79. Cambridge: Cambridge University Press. Diringer, David (1968) The Alphabet: a Key to the History of Mankind, Vol. 2, 3rd edn. Completely revised with the collaboration of Reinhold Regensburger. London: Hutchinson. Eco, Umberto (1976) A Theory of Semiotics. Bloomington: Indiana University Press. —— (1984) Semiotics and the Philosophy of Language. Bloomington: Indiana University Press. Evans, Jonathan D. and Andre Helbo (1986) Semiotics and International Scholarship: Towards a Language of Theory. Dordrecht: M. Nijhoff. Fairley, Irene R. (1975) E. E. Cummings and Ungrammar: a Study of Syntactic Deviance in His Poems. New York: Watermill Publishers. Fauconnier, Gilles and Mark Turner (1998) ‘Principles of Conceptual Integration’. In Discourse and Cognition, ed. Jean-Pierre Koenig, 269–83. Stanford, Calif.: CSLI Publications. Fauconnier, Gilles and Mark Turner (2002) The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic Books. Feibleman, James K. (1946) An Introduction to the Philosophy of C. S. Peirce. Cambridge, Mass.: MIT Press. Finegan, Edward (1994) Language: Its Structure and Use, 2nd edn. Orlando, Fla.: Harcourt Brace & Co. Fish, Stanley E. (1996 [1973]) ‘What is Stylistics and Why are They Saying Such Terrible Things about It?’ In The Stylistics Reader: From Roman Jakobson to the Present, ed. Jean Jacques Weber, 94–116. London: Arnold. Fleischman, Suzanne (1989) ‘Temporal Distance: a Basic Linguistic Metaphor’. Studies in Language, 13 (1): 1–50. Fónagy, Ivan (1963) Die Metaphern in der Phonetik. The Hague: Mouton. —— (1965) ‘Form and Function in Poetic Language’. Diogenes, 51: 71–110. —— (1979) La métaphore en phonétique. Montreal: Didier. —— (1999) ‘Why Iconicity?’ In Form Miming Meaning: Iconicity in Language and Literature, eds Max Nänny and Olga Fisher, 3–36. Amsterdam: John Benjamins. Freeman, Donald C. (1978a) ‘Syntax and Romantic Poetics’. In Proceedings of the 12th International Congress of Linguists, eds Wolfgang U. Dressler and Wolfgang Meid, 654–7. Innsbrucker Beiträge zur Sprachwissenschaft. —— (1978b) ‘Keat’s “To Autumn”: Poetry as Process and Pattern’. Language and Style, 11: 3–17. —— (1993) ‘ “According to My Bond”: King Lear and Re-Cognition’. Language and Literature, 2(1): 1–18. —— (1995) ‘ “Catch[ing] the Nearest Way”: Macbeth and Cognitive Metaphor’. Journal of Pragmatics, 24(6): 689–708. Freeman, Margaret (1995) ‘Metaphor Makes Meaning: Dickenson’s Conceptual Universe’. Journal of Pragmatics, 24(6): 643–66. —— (1997) ‘Grounded Spaces: Deictic-Self Anaphors in the Poetry of Emily Dickinson’. Language and Literature, 6(1): 7–28. —— (2000) ‘Poetry and the Scope of Metaphor: Toward a Cognitive Theory of Literature’. In Metaphor and Metonymy at the Crossroads, ed. Antonio Barcelona, 253–80. New York: Mouton de Gruyter.

242 References Friedrich, Paul (1986) The Language Parallax: Linguistic Relativism and Poetic Indeterminacy. Austin, Tex.: University of Texas Press. Fromkin, Victoria and Robert Rodman (1983) Introduction to Language, 3rd edn. New York: CBS College Publishing. Fujii, Yoko (1986) ‘A Comparative Study of “Freeze” in Japanese and English’. In Cross-Cultural Communication: East and West, Vol. 1, eds John H. Koo and Robert N. St. Clair, 335–55. Tainan: T’ai Ch’eng Publishing. [Distributed by the Department of Foreign Languages and Literature, National Cheng-Kung University, Tainan, Taiwan.] Gavins, Joanna and Gerard Steen (eds) (2003) Cognitive Poetics in Practice. London: Routledge. Gelb, Ignace J. (1963) A Study of Writing. Chicago: University of Chicago Press. Gibbs, Raymond W. Jr (1994) The Poetics of Mind. Cambridge: Cambridge University Press. Givón, Talmy (1985) ‘Iconicity, Isomorphism and Non-Arbitrary Coding in Syntax’. In Iconicity in Syntax, ed. John Haiman, 187–219. Amsterdam: John Benjamins. Goody, Jack (1987) The Interface between the Written and the Oral. Cambridge: Cambridge University Press. Group m (1970) Rhétorique générale. Paris: Larousse. Haberland, Hartmut (1994) ‘Written and Spoken Language: Relationship’. In The Encyclopedia of Language and Linguistics, Vol. 9, ed. R. E. Asher, 5060–2. Oxford: Pergamon. Haga, Jun (1971) ‘Nihongo Moji no Ninchi ni Kannsuru Ichi Kenkyu ‘[A Study on the Cognition of Japanese Letter-Types]. Keiryoo Kokugogaku [Mathematical Linguistics] 57: 29–34. Haiman, John (1985a) Natural Syntax. Amsterdam: John Benjamins. —— (ed.) (1985b) Iconicity in Syntax. Amsterdam: John Benjamins. Haley, Michael C. (1988) The Semeiosis of Poetic Metaphor. Bloomington, Ind.: Indiana University Press. Halliday, Michael A. K. (1978) Language as Semiotic. London: Edward Arnold. —— (1994a) An Introduction to Functional Grammar, 2nd edn. London: Edward Arnold. —— (1994b) ‘The Construction of Knowledge and Value in the Grammar of Scientiﬁc Discourse, with Reference to Charles Darwin’s The Origin of Species’. In Advances in Written Text Analysis, ed. Malcolm Coulthard, 136–56. London: Routledge. Halphern, Jack (ed.) (1993) NTC’s New Japanese–English Character Dictionary. Lincolnwood, Ill.: NTC Publishing Group. Hamano, Shoko Saito (1986) ‘The Sound-Symbolic System of Japanese’. PhD dissertation, The University of Florida. Ann Arbor: UMI Dissertation Information Service. Hamilton, Craig (2003) ‘A Cognitive Grammar of “Hospital Barge” by Wilfred Owen’. In Cognitive Poetics in Practice, eds Joanna Gavins and Gerard Steen, 55–65. London: Routledge. Harada, Shin-Ichi (1976) ‘Honoriﬁcs’. In Japanese Generative Grammar, ed. Masayoshi Shibatani, 499–561. New York: Academic Press. Harris, Roy (1984) ‘The Semiology of Textualization’. Language Sciences, 6(2): 271–86.

References 243 —— (1986) The Origin of Writing. London: Duckworth. —— (1989) ‘How Does Writing Restructure Thought?’ Language and Communication, 9(2/3): 99–106. —— (1995) The Signs of Writing. London: Routledge. —— (1996) Signs, Language, and Communication. London: Routledge. —— (1998a) ‘The Integrationist Critique of Orthodox Linguistics’. In Integrational Linguistics: a First Reader, eds Roy Harris and George Wolf, 15–26. Oxford: Pergamon. —— (1998b) Introduction to Integrational Linguistics. Oxford: Pergamon. Hastings, Robert (1963) ‘ “Easter Wings” as Model of Herbert’s Method’. Thoth, 4: 15–23. Hattori, Doho (1973 [1776]) ‘Sanzooshi’ [Three Notebooks]. In Rengaronshu, Nogakuronshu, Haironshu [Essays on Linked Poetry, Noh Play, and Haiku], eds and annot. Tetsuo Ichiji et al., 517–624. Tokyo: Shogakkan. Havelock, Eric A. (1976) Origins of Western Literacy. Toronto: Ontario Institute for Studies in Education. —— (1991 [1989]) ‘The Oral–Literal Equation: a Formula for the Modern Mind’. In Literacy and Orality, eds David R. Olson and Nancy Torrance, 11–27. Cambridge: Cambridge University Press. Henderson, Harold G. (1958) An Introduction to Haiku. New York: Doubleday. Henle, Paul (1958) Language, Thought, and Culture. Ann Arbor, Mich.: University of Michigan Press. Herbert, George (1880 [1633]) ‘Easter Wings’. In The Temple. Cambridge: T. Buck and R. Daniel, Printers to the University of Cambridge. —— (1994 [1633]) ‘Easter Wings’. In The Works of George Herbert, 35. Ware, Hertfordshire: Wordsworth Editions. Higgins, Dick (1977) George Herbert’s Pattern Poems in Their Tradition. Unpublished Editions. West Glover, Vermont and New York. Hinton, Leanne, Johanna Nichols and John Ohala (eds) (1994) Sound Symbolism. Cambridge: Cambridge University Press. Hiraga, Masako K. (1986) ‘Suikoo no Kigooron: Basho no Semi no Ku no Suikoo ni tsuite’ [Semiotics of Revision: a Study of Revisions of Basho’s Haiku about the Cicada]. Kigoogaku Kenkyuu [Studia Semiotica: Journal of the Japanese Association for Semiotic Studies], 6: 127–38. —— (1987a) ‘External Stillness: a Linguistic Journey to Basho’s Haiku about the Cicada’. Poetics Today, 8(1): 5–18. —— (1987b) ‘The Hidden Melodies: a Linguistic Analysis of a Japanese Short Poem’. In the Proceedings of the XIV International Congress of Linguists, eds Werner Bahner, Joachim Schildt and Dieter Viehweger, 1945–9. Berlin: Akademie-Verlag. —— (1988) ‘Tanka no Poetics: Tomonori no Sakura no Uta ni kansuru Koozooteki Bunseki o chuushin ni’ [Poetics of Tanka: Structural Analysis of Tomonori’s Tanka about Cherry Blossoms]. Kigoogaku Kenkyuu [Studia Semiotica: Journal of the Japanese Association for Semiotic Studies], 8: 141–53. —— (1990) ‘Sound as Meaning: Iconicity in Edgar Allan Poe’s “The Bells” ’. Journal of the University of the Air, 8: 1–23. —— (1993) ‘Iconic Meanings of Visual Repetition in Poetry’. In Semiotics 1991, eds John Deely and Terry Prewitt, 95–103. New York: University Press of America.

244 References —— (1994) ‘Diagrams and Metaphors: Iconic Aspects of Language’. Journal of Pragmatics, 22(1): 5–21. —— (1995a) ‘Vision as Meaning: Iconicity in Basho’s Haiku in Oku no Hosomichi’. Journal of the University of the Air, 13: 109–25. —— (1995b) ‘Inyu ni Mirareru Gengo-kan ni tsuite’ [Japanese Metaphors for Language]. In Ningen to Kotoba [Man and Speech Act in Honor of IHL Director Yasuo Isami], ed. The Editorial Committee of the Isami Festschrift, 257–69. Tokyo: Libre Press. —— (1999a) ‘ “Blending” and an Interpretation of Haiku: a Cognitive Approach’. Poetics Today, 20(3): 461–81. —— (1999b) ‘Rough Sea and the Milky Way: “Blending” in a Haiku Text’. In Computation for Metaphors, Analogy and Agents, ed. Chrystopher Nehaniv, 27–36. Berlin: Springer-Verlag. —— (1999c) ‘DEFERENCE as DISTANCE: Metaphorical Base of Honoriﬁc Verb Construction in Japanese’. In Cultural, Psychological and Typological Issues in Cognitive Linguistics, eds Masako K. Hiraga, Chris Sinha and Sherman Wilcox, 47–68. Amsterdam: John Benjamins. —— (2000a) ‘The Interplay of Metaphor and Iconicity: a Cognitive Approach’. Unpublished Doctoral Dissertation, University of London, 364 pp. —— (2000b) ‘Iconicity as Principle of Composition and Interpretation: a Case Study of Japanese Short Poems’. In Phonosymbolism and Poetic Language, ed. Patrizia Violi, 147–69. Turnhout, Belgium: Brepols Publishers. —— (2002) ‘The Interplay of Metaphor and Iconicity: a Cognitive Approach to Poetic Texts’. Special Issue on Metaphor, ed. Tomasz Komendzinski, Theoria et Historia Scientiarum: International Journal for Interdisciplinary Studies, VI/1, 179–244. —— (2003) ‘How Metaphor and Iconicity Are Entwined in Poetry: a Case in Haiku’. In From Sign to Signing, eds Wolfgang Muller and Olga Fisher, 317–35. Amsterdam: John Benjamins. Hiraga, Masako K. and Joanna Radwanska-Williams (eds) (1994) Metaphor and Iconicity in Language. Special Issue of Journal of Pragmatics, 22 (1). Hockett, Charles F. (1958) ‘Secondary Effects of Phonemic Shapes’. In A Course in Modern Linguistics, 293–300. New York: The Macmillan Co. Hodge, Robert and Gunter Kress (1988) Social Semiotics. Cambridge: Polity Press. Hollander, John (1991 [1969]) ‘Swan and Shadow’. In Types of Shape, 35. New Haven: Yale University Press. Hopper, Paul J. (1998) ‘Emergent Grammar’. In The New Psychology of Language: Cognitive and Functional Approaches to Language Structure, ed. Michael Tomasello, 155–75. Mahwah, NJ and London: Lawrence Erlbaum Associates. Hughes, Glenn (1960) Imagism and the Imagists: a Study in Modern Poetry. New York: The Humanities Press. Ide, Sachiko (1982) ‘Japanese Sociolinguistics: Politeness and Women’s Language’. Lingua, 57: 357–86. Ikegami, Yoshihiko (1996) ‘Some Traditional Japanese Visual Tropes and their Perceptual and Experiential Bases’. Poetica, 46: 89–99. Ikegami, Yoshihiko, Keiichi Yamanaka and Norimitsu Tohsu (1983) Bunka-KigooRon e-no Shootai [Invitation to Semiotics of Cultures]. Tokyo: Yuuhikaku. Imoto, Noichi (1978) Basho no Bungaku no Kenkyu [Study of Basho’s Literature]. Tokyo: Kadokawa Shoten.

References 245 Isami, Yasuo (1994 [1974]) ‘Seikoodoo toshite-no Kotoba’ [Speech as an Act of Life]. In Ningen to Kotoba [Man and Speech Act in Honor of IHL Director Yasuo Isami], ed. The Editorial Committee of the Isami Festschrift, 32–8. Tokyo: Libre Press. —— (1994 [1990]) ‘Ningen Juushi-no Gengo Kenkyuu e: Human Linguistics no Kadai’ [Toward a Study of Language with an Emphasis on Human Beings: Issues in Human Linguistics]. In Ningen to Kotoba [Man and Speech Act in Honor of IHL Director Yasuo Isami], ed. The Editorial Committee of the Isami Festschrift, 39–51. Tokyo: Libre Press. —— (1994) ‘Kanren toshite-no Hiyu’ [Metaphor as Relativity]. In Ningen to Kotoba [Man and Speech Act in Honor of IHL Director Yasuo Isami], ed. The Editorial Committee of the Isami Festschrift, 8–12. Tokyo: Libre Press. Ishikawa, Minako (1991) ‘Iconicity in Discourse: the Case of Repetition’. Text, 11(4): 553–80. Janney, Richard W. (1999) ‘Words as Gestures’. Journal of Pragmatics, 31(7): 953–72. Jakobson, Roman (1960) ‘Linguistics and Poetics’. In Style in Language, ed. Thomas A. Sebeok, 350–77. Cambridge, Mass.: MIT Press. —— (1971 [1965]) ‘Quest for the Essence of Language’. In Selected Writings II, 345–59. The Hague: Mouton. —— (1971) Selected Writings V. The Hague: Mouton. —— (1981 [1979]) ‘Notes on the Contours of an Ancient Japanese Poem: the Farewell Poem of 732 by Takapasi Musimaro’. In Selected Writings III, 157–64. The Hague: Mouton. —— (1981) Selected Writings III. The Hague: Mouton. —— (1985 [1960]) ‘Poetry of Grammar and Grammar of Poetry’. In Verbal Art, Verbal Sign, Verbal Time, eds Krystyna Pomorska and Stephen Rudy, 37–46. Minneapolis, Minn.: University of Minnesota Press. —— (1985 [1980]) ‘Subliminal Verbal Patterning in Poetry’. In Verbal Art, Verbal Sign, Verbal Time, eds Krystyna Pomorska and Stephen Rudy, 59–68. Minneapolis, Minn.: University of Minnesota Press. —— (1985) Verbal Art, Verbal Sign, Verbal Time, eds Krystyna Pomorska and Stephen Rudy. Minneapolis: University of Minnesota Press. Jakobson, Roman, C. Gunnar, M. Fant and Morris Halle (1951) Preliminaries to Speech Analysis. Cambridge, Mass.: MIT Press. Jakobson, Roman and Stephen Rudy (1985 [1980]) ‘Yeats’ “Sorrow of Love” through Years’. In Verbal Art, Verbal Sign, Verbal Time, eds K. Pomorska and Stephen Rudy, 79–101. Minneapolis, Minn.: University of Minnesota Press. Jakobson, Roman and Linda R. Waugh (1979) The Sound Shape of Language. Bloomington, Ind.: Indiana University Press. James, Vincent and Erin Jansen (1995–2000) The Internet Language Dictionary. NetLingo Inc. (http://www.netlingo.com). Jespersen, Otto (1933 [1922]) ‘Symbolic Value of the Vowel i’. Linguistica, 283–303. —— (1964 [1921]) ‘Sound Symbolism’. In Language: Its Nature, Development and Origin, 396–411. New York: Norton. Johnson, Mark (1987) The Body in the Mind: the Bodily Basis of Meaning, Imagination, and Reason. Chicago: University of Chicago Press. Jones, Peter (ed.) (1972) Imagist Poetry. Harmondsworth: Penguin.

246 References Kakehi, Hisao, Ikuhiro Tamori and Lawrence Schourup with the assistance of Leslie James Emerson (eds) (1996) Dictionary of Iconic Expressions in Japanese. Berlin: M. de Gruyter. Kaneko, Yoshio (1973) Oku no Hosomichi no Kenkyu [A Study of Oku no Hosomichi]. Tokyo: Ohfusha. Kawamoto, Koji (1991) Nippon Shiika no Dentoo: 7 to 5 no Shigaku [Tradition of Japanese Poetry: Poetics of 7 and 5]. Tokyo: Iwanami Shoten. Kawamoto, Shigeo (1978) Kotoba no Shikisai [Colours of Words]. Tokyo: Iwanami Shoten. —— (1986) Kotoba to Imeji [Words and Images]. Tokyo: Iwanami Shoten. Kennedy, X. J. (1986) An Introduction to Poetry, 6th edn. Boston: Little, Brown, & Co. Kikuchi, Shigeo (1984) ‘How Autumn Comes: a Jakobsonian Approach to Waka’. Studies in English Linguistics and Literature, 1(1): 5–12. Kikuchi, Yasuto (1994) Keigo. [Honoriﬁcs]. Tokyo: Kadokawa. Kino, Tomonori (1971 [905]) ‘Hisakata no’ [Tanka on Cherry Blossoms]. In Kokin Waka-shu [Imperial Anthology of Waka], ed. Tsurayuki Kino et al. and annot. Masao Ozawa, 88–9. Tokyo: Shoogakkan. Kino, Tsurayuki (1971 [905]) ‘Kanajo’ [Introduction Written in Kana]. In Kokin Waka-shu [Imperial Anthology of Waka], ed. Tsurayuki Kino et al. and annot. Masao Ozawa, 42–62. Tokyo: Shoogakkan. Kobayashi, Shojiro (1966) ‘Basho to Kireji’ [Basho and Cutting Letters]. Gengo to Bungei [Language and Literature], 47. Kress, Gunther and Theo van Leeuwen (1996) Reading Images: the Grammar of Visual Design. London: Routledge. Kuno, Susumu (1973) The Structure of the Japanese Language. Cambridge, Mass.: MIT Press. Ladefoged, Peter (1982) A Course in Phonetics, 2nd edn. New York: Harcourt Brace Jovanovich. Lakoff, George (1987) Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: Chicago University Press. —— (1993) ‘The Contemporary Theory of Metaphor’. In Metaphor and Thought, ed. Andrew Ortony, 202–51. Cambridge: Cambridge University Press. Lakoff, George and Mark Johnson (1980) Metaphors We Live By. Chicago: Chicago University Press. Lakoff, George and Mark Turner (1989) More than Cool Reason: a Field Guide to Poetic Metaphor. Chicago: Chicago University Press. Landsberg, Marge E. (1979) ‘The Fair Moors of Oceania’. Mankind Quarterly, 19(3): 249–64. —— (1980) ‘The Icon in Semiotic Theory’. Current Anthropology, 21(1): 93–5. —— (ed.) (1995) Syntactic Iconicity and Linguistic Freezes: the Human Dimension. Berlin: Mouton de Gruyter. Langacker, Ronald W. (1987) Foundations of Cognitive Grammar, Vol. 1. Stanford: Stanford University Press. —— (1998) ‘Conceptualization, Symbolization, and Grammar’. In The New Psychology of Language: Cognitive and Functional Approaches to Language Structure, ed. Michael Tomasello, 1–39. Mahwah, NJ and London: Lawrence Erlbaum Associates.

References 247 Langer, Susanne K. (1957) Philosophy in a New Key: a Study in the Symbolism of Reason, Rite, and Art. Cambridge, Mass.: Harvard University Press. Lee, David (2001) Cognitive Linguistics: Introduction. Oxford: Oxford University. Leech, Geoffrey N. (1969) A Linguistic Guide to English Poetry. London: Longman. Lennard, John (1991) But I Digress: the Exploitation of Parentheses in English Printed Verse. Oxford: Clarendon Press. Lyons, John (1972) ‘Human Language’. In Non-Verbal Communication, ed. Robert A. Hinde, 49–85. Cambridge: Cambridge University Press. Ma, Kang Yuen (1997) ‘The Relation between the Writing System and the Use of Metaphor in English and Chinese’. Unpublished MA thesis, University of North Dakota. McGough, Roger (1989 [1971]) ‘40–Love’. In Selected Poems 1967–1987, 84–5. London: Jonathan Cape. McLuhan, Marshall (1994 [1964]) Understanding Media: the Extensions of Man. Cambridge, Mass.: MIT Press. McNeill, David (1992) Hand and Mind: What Gestures Reveal about Thought. Chicago: The University of Chicago Press. Matsuo, Basho (1957 [1694]) Oku no Hosomichi (Sora Zuikoo Nikki tsuki) [The Narrow Road to the Deep North and Sora’s Travel Diary]. Annot. Shoichiro Sugiura. Tokyo: Iwanami Shoten. —— (1966 [1694]) ‘The Narrow Road to the Deep North’. In The Narrow Road to the Deep North and Other Travel Sketches, trans. and ed. Nobuyuki Yuasa, 97–143. London: Penguin Books. —— (1972) Matsuo Basho Shu [Works of Matsuo Basho]. Eds and annot. Noichi Imoto, et al. Tokyo: Shogakkan. —— (1985) Basho Buson Hokku Soosakuin [Complete Collection of Hokku by Basho and Buson]. Eds Takehiko Michimoto and Yoskazu Taniguchi, 2 vols. Tokyo: Kadokawa Shoten. —— (1996 [1694]) Basho’s Narrow Road. Trans. Hiroaki Sato. Berkeley, Calif.: Stonebridge Press. —— (1997 [1690]) Basho Jihitsu Oku no Hosomichi [Oku no Hosomichi by Basho’s Handwriting]. Eds Yozo Ueno and Takejiro Sakurai. Tokyo: Iwanami Shoten. Miron, Murray, S. (1961) ‘A Cross-Linguistic Investigation of Phonetic Symbolism’. Journal of Abnormal and Social Psychology, 62: 623–30. Morita, Ran (1970) Basho no Hoohoo [Method of Basho]. Tokyo: Kyooiku Shuppan Center. Morton, J. and S. Sasanuma (1984) ‘Lexical Access in Japanese’. In Orthographies and Reading: Perspectives from Cognitive Psychology, Neuropsychology and Linguistics, ed. Leslie Henderson, 25–42. London: Lawrence Erlbaum Associates. Müller, Wolfgang and Olga Fischer (eds) (2002) From Sign to Signing: Iconicity in Language and Literature 3. Amsterdam: John Benjamins. Muramatsu, Tomotsugu (1988) Sora-Bon Oku no Hosomichi no Kenkyu [A Study of Sora’s Copy of Oku no Hosomichi]. Tokyo: Kasama Shoin. Nänny, Max and Olga Fischer (eds) (1999) Form Miming Meaning: Iconicity in Language and Literature. Amsterdam: John Benjamins. Nihon Daijiten Kankookai (1976) Nihon Kokugo Dai-Jiten. [Dictionary of the Japanese Language]. Tokyo: Shogakkan.

248 References Niikuni, Seiichi (1985 [1966]) ‘Ame’ [Rain]. Typographics ‘TEE’, June: 21. Nishida, Naotoshi (1987) Keigo. [Honoriﬁcs]. Tokyo: Tokyo-doo. Nomura, Masuhiro (1996) ‘The Ubiquity of the Fluid Metaphor in Japanese: a Case Study’. Poetica, 46. 41–75. Norrick, Neal R. (1987) ‘Function of Repetition in Conversation’. Text, 7(3): 245–64. Ogata, Tsutomu (1971) Matsuo Basho. Tokyo: Chikuma Shoboo. Ohno, Susumu (1958) Kana-Moji/Kana-Bun no Sooshi [Creation of Kana Characters and Kana Literature]. Tokyo: Iwanami Shoten. Ohori, Toshio (1987) ‘On the Concept of Iconicity in the Study of Language’. Language, Culture and Communication, 2: 34–50. Keio University. —— (1991) ‘Bunpoo Koozoo no Ruizoosei’ [Iconicity in Grammatical Structures]. Studia Semiotica: The Journal of the Japanese Association for Semiotic Studies, 11: 95–107. Ohtani, Tokuzo (ed.) (1962) Basho Ku-Shu [Collection of Basho’s Haiku]. Tokyo: Iwanami Shoten. Olson, David R. (1994) The World on Paper. Cambridge: Cambridge University Press. Olson, David R. and Nancy Torrance (eds) (1991) Literacy and Orality. Cambridge: Cambridge University Press. Ong, Walter (1982) Orality and Literacy. London: Methuen. Ozawa, Masao (1971 [905]) Kokin Waka-Shu [Imperial Anthology of Waka]. Tokyo: Shoogakkan. Paradis, Michel, Hiroko Hagiwara and Nancy Hildebrandt (1985) Neurolinguistic Aspects of the Japanese Writing System. Tokyo: Academic Press Japan. Peer, Willie van (1993) ‘Typographic Foregrounding’. Language and Literature, 2(1): 49–61. Peirce, Charles S. (1931–58) Collected Papers of Charles Sanders Peirce. Eds Charles Hartshorne and Paul Weiss. Cambridge: Harvard University. —— (1962 [1955, 1902]) ‘Logic and Semiotic: Theory of Signs’. In Philosophical Writings, ed. Justus Buchler, 98–119. New York: Dover. Plato (1926) ‘Cratylus’. In Plato, Vol. IV. Trans. H. N. Fowler. London: William Heinemann. Poe, Edgar Allan (1919) ‘The Bells’. In The Complete Poems of Edgar Allan Poe, ed. James H. Whitty, 63–6. Boston: Houghton Mifﬂin. Posner, Roland (1986) ‘Iconicity in Syntax’. In Iconicity: Essays on the Nature of Culture, eds Paul Bouissac et al., 305–37. Tübingen: Stauffenburg. Poster, Mark (1990) The Mode of Information. Cambridge: Polity Press. Pound, Ezra (1928 [1916]) ‘In a Station of the Metro’. In Selected Poems, 53. London: Faber and Faber. —— (1936) The Chinese Written Characters as a Medium for Poetry. San Francisco: City Light Books. Reddy, Michael (1979) ‘The Conduit Metaphor’. In Metaphor and Thought, ed. Andrew Ortony, 284–324. Cambridge University Press. Richards, I. A. (1936) The Philosophy of Rhetoric. Oxford University Press. Ricoeur, Paul (1977) The Rule of Metaphor: Multi-Disciplinary Studies of the Creation of Meaning in Language. Trans. Robert Czerny with Kathleen McLaughlin and John Costello. Toronto: University of Toronto Press. —— (1978) The Philosophy of Paul Ricoeur. Boston: Beacon Press.

References 249 —— (1979 [1978]) ‘Metaphorical Process as Cognition, Imagination, and Feeling’. In On Metaphor, ed. Sheldon Sacks, 141–57. Chicago: University of Chicago Press. —— (1981) Hermeneutics and the Human Science. Cambridge: Cambridge University Press. Ross, John Robert (1981) ‘Robert Frost’s “Out, Out –” ’. In Crossing the Boundaries in Linguistics, eds Wolfgang Klein and Willem Levelt, 265–82. Dordrecht: D. Reidel Publishing Co. —— (1982a) ‘Hologramming in a Robert Frost Poem: the Still Point’. In Linguistics in the Morning Calm, 685–91. Seoul: Hanshin Publishing Co. —— (1982b) ‘Poems as Holograms’. Poetics Journal, 3–11. East Haven, Conn.: Inland Book Co. —— (1982c) ‘That Depends’. Unpublished MS. —— (1986) ‘Languages as Poems’. In The Interdependence of Theory, Data, and Application: Georgetown University Round Table 85, ed. Deborah Tannen, 180–204. Washington, DC.: Georgetown University Press. —— (1989) ‘Visible Rhythm’. Unpublished MS. Universidade Federal de Minas Gerais. —— (1990) ‘The Curve of Love’. In Linguistic Fiesta: Festschrift for Professor Hisao Kakehi’s Sixtieth Birthday, 151–82. Tokyo: Kuroshio Shuppan. —— (2000) ‘The Taoing of a Sound: Phonetic Drama in William Blake’s The Tyger’. In Phonosymbolism and Poetic Language, ed. Patrizia Violi, 99–145. Turnhout, Belgium: Brepols. Sampson, Geoffrey (1985) Writing Systems. Stanford, Calif.: Stanford University Press. Sapir, Edward (1921) Language. New York: Harcourt, Brace and World. Sarup, Madan (1993) An Introductory Guide to Post-Structuralism and Postmodernism, 2nd edn. New York: Harvester Wheatsheaf. Saussure, Ferdinand de (1983 [1916]) Course in General Linguistics. Trans. Roy Harris. London: Duckworth. Semino, Elena and Jonathan Culpeper (eds) (2002) Cognitive Stylistics: Language and Cognition in Text Analysis. Amsterdam: John Benjamins. Shapiro, Karl and Robert Beum (1965) A Prosody Handbook. New York: Harper and Row. Shapiro, Michael (1976) Asymmetry: an Inquiry into the Linguistic Structure of Poetry. Amsterdam: North-Holland. —— (1998) ‘Sound and Meaning in Shakespeare’s Sonnets’. Language, 74(1): 81–103. Shelley, Percy Bysshe (1904 [1820]) ‘Love’s Philosophy’. In The Golden Treasury, ed. Francis T. Palgrave, 216. London: Macmillan – now Palgrave Macmillan. —— (1943) The Complete Poetical Works of Percy Bysshe Shelley. Ed. Thomas Hutchinson. Oxford: Oxford University Press. Shibasaki, Reijirou (2001) ‘A Blending Analysis of Chinese and Japanese Homographic Compounds: Toward a Cognitive Account of Visual Iconicity’. In the Proceedings of the 1st Seoul International Conference on Discourse and Cognitive Linguistics, 267–80. Seoul, Korea. Shibatani, Masayoshi (1990) The Languages of Japan. Cambridge: Cambridge University Press. Shirane, Haruo (1998) Traces of Dreams: Landscape, Cultural Memory, and the Poetry of Basho. Stanford: Stanford University Press.

250 References Simone, Raffaele (ed.) (1995) Iconicity in Language. Amsterdam: John Benjamins. Solt, Mary Ellen (ed.) (1970) Concrete Poetry: a World View. Bloomington: Indiana University Press. Sorensen, Heidi (1999) ‘Communication in Electronic Graphic Space’. Unpublished MA thesis, Goldsmiths College, University of London. Steen, Gerard (2003) ‘ “Love Stories”: Cognitive Scenarios in Love Poetry’. In Cognitive Poetics in Practice, eds Joanna Gavins and Gerard Steen, 67–82. London: Routledge. Stockwell, Peter (2002) Cognitive Poetics: an Introduction. London: Routledge. Summers, Joseph (1954) George Herbert: His Religion and Art. London: Chatto and Windus. Sweetser, Eve (1990) From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press. —— (1995) ‘Metaphor, Mythology, and Everyday Langauge’. Journal of Pragmatics, 24(6): 585–93. Tai, James H-Y (1985) ‘Temporal Sequence and Chinese Word Order’. In Iconicity in Syntax, ed. John Haiman, 49–72. Amsterdam: John Benjamins. Talmy, Leonard (1978) ‘Figure and Ground in Complex Sentences’. In Universals of Human Language, Vol. 4, ed. Joseph H. Greenberg, 625–49. Stanford: Stanford University Press. Tannen, Deborah (1987) ‘Repetition in Conversation as Spontaneous Formulaicity’. Text, 7(3): 215–43. —— (1990) ‘Ordinary Conversation and Literary Discourse: Coherence and the Poetics of Repetition’. In The Uses of Linguistics: Annals of the New York Academy of Sciences, Vol. 583, ed. Edward H. Bendix, 15–32. Tanz, Christine (1971) ‘Sound Symbolism in Words Relating to Proximity and Distance’. Language and Speech, 14: 266–76. Taub, Sarah (2001) Language from the Body: Iconicity and Metaphor in American Sign Language. Cambridge: Cambridge University Press. Toolan, Michael (1996) Total Speech: an Integrational Linguistic Approach to Language. Durham: Duke University Press. Toyama, Shigehiko (1970) ‘Shooryaku no Bungaku: Kireji-Ron’ [Literature of Ellipsis: Theory of Cutting Letters]. In Hassoo to Hyoogen [Conception and Expression], ed. Michiyoshi Kadokawa, 320–50. Tokyo: Kadokawa Shoten. Tsujimura, Toshiki (1968) Keigo no Shiteki Kenkyuu [Historical Study of Honoriﬁcs]. Tokyo: Tokyo-doo. Tsur, Reuven (1983) ‘What is Cognitive Poetics?’ Papers in Cognitive Poetics, 1. The Katz Institute for Hebrew Literature, Tel Aviv University. —— (1992a) Toward a Theory of Cognitive Poetics. Amsterdam: Elsevier (North-Holland). —— (1992b) What Makes Sound Patterns Expressive: the Poetic Mode of Speech Perception. Durham: Duke University Press. —— (1996) ‘Rhyme and Cognitive Poetics’. Poetics Today, 17(1): 55–84. Turner, Mark (1987) Death is the Mother of Beauty: Mind, Metaphor, Criticism. Chicago: University of Chicago Press. —— (1991) Reading Minds. Princeton: Princeton University Press. —— (1996) The Literary Mind. Oxford: Oxford University Press. —— (1998) ‘Figure’. In Figurative Language and Thought, eds Albert Katz, Cristina Cacciar, Raymond W. Gibbs, Jr. and Mark Turner, 44–87. Oxford: Oxford University Press.

References 251 Turner, Mark and Gilles Fauconnier (1995a) ‘Conceptual Integration and Formal Expression’. Metaphor and Symbolic Activity, 10(3): 183–204. —— (1995b) ‘Blending and Metaphor’. Unpublished MS. —— (1999) ‘Mechanism of Creativity’. Poetics Today, 20(3): 397–418. Tuve, Rosemond (1952) A Reading of George Herbert. Chicago: University of Chicago Press. Ueno, Yozo (1986) Basho-Ron [On Basho]. Tokyo: Chikuma Shobo. Ultan, Russell (1978) ‘Size-Sound Symbolism’. In Universals of Human Language, ed. Joseph H. Greenburg, 525–68. Stanford: Stanford University Press. Umegaki, Minoru (1974) ‘Hoogen Keigo Kokoroechoo’ [Notes on Honoriﬁcs in Dialects]. In Keigo Yoohoo Jiten [Dictionary of Honoriﬁc Usage], 155–96. Tokyo: Meiji-shoin. Ungerer, Friedrich (1999) ‘Diagrammatic Iconicity in Word-Formation’. In Form Miming Meaning: Iconicity in Language and Literature, eds Max Nänny and Olga Fisher, 307–24. Amsterdam: John Benjamins. Ungerer, Friedrich and Hans-Jorg Schmid (1996) An Introduction to Cognitive Linguistics. London, New York: Longman. Vance, Timothy J. (1987) An Introduction to Japanese Phonology. New York: State University of New York Press. Walker, John David (1962) ‘The Architectonics of George Herbert’s The Temple’. English Literary History, 29, 289–306. Watanabe, Shooichi et al., trans. (1986) Retorikku to Jinsei [Metaphors We Live By]. Tokyo: Taishuukan. Waugh, Linda R. (1992) ‘Presidential Address: Let’s Take the Con out of the Iconicity: Constrains on Iconicity in the Lexicon’. American Journal of Semiotics, 9: 7–48. —— (1994) ‘Degrees of Iconicity in Lexicon’. Journal of Pragmatics, 22: 55–75. Waugh, Linda R. and Madeleine Newﬁeld (1995) ‘Iconicity in the Lexicon and Its Relevance for a Theory of Morphology’. In Syntactic Iconicity and Linguistic Freezes: the Human Dimension, ed. Marge E. Landsberg, 189–221. Berlin: Mouton de Gruyter. Weber, Jean Jacques (ed.) (1996) The Stylistics Reader: From Roman Jakobson to the Present. London: Arnold. Weiss, Jonathan, H. (1963) ‘Role of “Meaningfulness” Versus Meaning Dimensions in Guessing the Meanings of Foreign Words’. Journal of Abnormal and Social Psychology, 66: 541–6. Wellek, Rene and Austin Warren (1963 [1949]) Theory of Literature, 3rd edn. Harmondsworth, Middlesex: Penguin Books. Wescott, Roger Williams (1971) ‘Linguistic Iconism’. Language, 47(2): 417– 28. Westerweel, Bart (1984) Patterns and Patterning: a Study of Four Poems by George Herbert. Amsterdam: Rodopi. Wildman, Eugene (ed.) (1969) Anthology of Concretism. Chicago: Swallow Press. Wilkinson, Valerie Anne (1997) ‘Allegorical Exegesis: a New Approach’. Studies in Information, 2: 53–60, Shizuoka University. Williams, Emmett (ed.) (1967) An Anthology of Concrete Poetry. New York: Something Else Press. Williams, William Carlos (1970 [1923]) ‘Spring and All’. In William Carlos Williams: Imaginations, ed. Webster Schott, 88–151. New York: A New Directions Book.

252 References Wordsworth, William (1983 [1804]) ‘My Heart Leaps up When I Behold’. In Poems in Two Volumes, and Other Poems, 1800–7. Ithaca, NY: Cornell University Press. Worth, Dean S. (1977) ‘Roman Jakobson and the Study of Rhyme’. In Roman Jakobson: Echoes of His Scholarship, eds Daniel Armstrong and C. H. van Schooneveld, 515–33. Lisse: The Peter de Ridder Press. Wyss, Eva Lia (1999) ‘Iconicity in the Digital World’. In Form Miming Meaning: Iconicity in Language and Literature, eds Max Nänny and Olga Fisher, 285–304. Amsterdam: John Benjamins. Yamamoto, Kenkichi (1974) Basho Zen-Hokku-Shu [Complete Collection of Basho’s Hokku]. Tokyo: Kawade Shoboo Shinsha. —— (1984 [1951]) ‘Aisatsu to Kokkei’ [Greeting and Humour]. In Yamamoto Kenkichi Zenshuu [Collected Writings of Kenkichi Yamamoto], Vol. 8, 21–66. Tokyo: Koodansha. Yamamura, Bocho (1989 [1915]) ‘Fuukei’ [Scene]. In Yamamura Bocho Zenshuu [Collection of Poems by Bocho Yamamura], 89–90. Tokyo: Chikuma Shoboo. Yamanashi, Masa-aki (1974) ‘On Minding Your p’s and q’s in Japanese: a Case Study from Honoriﬁcs’. In Papers from the 10th Regional Meeting, Chicago Linguistic Society, eds Michael W. La Galy, Robert A. Fox and Anthony Bruck. Yasuda, Kenneth (1957) The Japanese Haiku. Rutland, Vt.: Charles E. Tuttle Co. Yule, George (1985) The Study of Language: an Introduction. Cambridge: Cambridge University Press.

Index Akabane, M. 17 Akimoto, F. 17 allegorical interpretation 33, 74, 75, 78, 82 alliteration 143, 145 alphabet and the primacy of speech 195 vs. non-alphabetic writing systems 195–7 alphabetic symbols 93 alphabetic system 196, 209 ‘Ame (rain)’ 111–12 analogy 32, 34, 35, 39, 44 Anderson, D. 15, 35 arbitrariness 14–15 arbitrary 3, 14 Arima, M. 17 Aristotle 23 Asano, T. 234 n. 6 Aso, I. 17 Attridge, D. 23 auditory representation, in poetry 127–34 Ayer, A. 29 Beardsley, M. 24 Becker, A. L. 224 ‘The Bells’ 134, 135–46 alliteration 145 diagrammatic iconicity 142–6 imagic iconicity 139–42 metaphor–image links 139 Bennett, J. 230 n. 1 Birdwhistell, R. L. 226 Black, M. 24 Blake, William 91 blank space 118 Blasko, D. G. and D. W. Merski 229 n. 17 blending model 16, 22, 36–41, 52, 89, 222–3 basic claims 36–8 and haiku 71

and kanji 205–6 metaphor–icon links 40 Bloomﬁeld, L. 178, 195, 208 Blyth, R. H. 228 n. 8 Bolinger, D. L. 24, 43, 176, 177, 208, 237 n. 1 Bolter, D. J. 212 Bouissac, P. et al. 28 Boultenhouse, C. 230 n. 1 Brogan, T. 106, 209 Brooks, C. and R. P. Warren 18 Brown, C. C. and W. P. Ingoldsby 230 n. 1 Brown, P. and S. Levinson 186, 191 Brown, R. and A. Gilman 186 Brown, R. and R. Nattal 209 Brown, R. W. et al. 209 Bybee, J. L. 180 calligraphy 238 n. 7 capitalised words 118 Chinese characters as ideograms 197–9 logographic nature of 197–8 Claudi, U. and B. Heine 184 cognitive account, of poetic metaphor 26–7 cognitive approach, to metaphor–icon links 15–17, 19 cognitive deﬁnition 35–45 cognitive linguistics 16, 25–8, 229 n. 15 cognitive metaphors and haiku 70–2 ‘Love’s Philosophy’ 64 cognitive poetics 26–7, 229 n. 3 cognitive processes 13 cognitive role, of written language 19 cognitive theory, of metaphor 25–6 cognitive treatment, of iconicity 27–8 253

254 Index communication by computer 226 non-verbal 226–7 communication media, ecology of 225–7 competing motivations 192–3 compound logographs 200, 201–3 computer communication by 226 mind conceptualised as 5–6, 33, 38–9, 228 n. 5 computers 211–12 concrete poetry 106–15, 116 metaphor–icon links in 107–8 conduit metaphor 41, 174–5, 230 n. 9 consonants 164 sound-symbolism of 132–3 contiguity 30 conventionality 9, 10, 13, 14, 27, 30, 33 Cooper, W. E. and J. R. Ross 179–80 Course in General Linguistics 14 Cratylus 14 Crystal, D. 232 n. 2 cultural cognitive model, and haiku 72 Cummings, E. E. ‘1(a’ 109–11, 233 n. 8 Danesi, M. 6, 25, 229 n. 14 Davis and Taylor 225 Deacon, T. W. 230 n. 5 Deane, P. 26, 27 deference 189, 192 DeFrances, J. 198 Derrida, J. 209 Deuchar, M. 17 diacritics 232 n. 5 diagrammatic content, metaphor 6 diagrammatic correlation 42–3 diagrammatic manifestations, visual poetry 106–15 diagrammatic structuring, of sound 156–68 diagrams 32, 34 structural and relational 43–5 Dickinson, Emily 27 Diringer, D. 196

discourse, iconicity in 185–6 discursive and presentational representation 18–19, 213–15 distance 186–92 horizontal 191 temporal 191–2 dynamic nature of language 224 ‘Easter Wings’ 104, 230 n. 1 global metaphor in 59, 61 grammatical metaphors 61–2 imagic iconicity by visual layout 62–3 metaphor–icon links in 60 space and shape in 58–63 Eco, Umberto 28 elaborated grammatical metaphors 175 electronic communication, iconicity in 217–18 ‘emergent grammar’ 225 emoticons 197, 215–16 English sound system, compared to Japanese 128–31 English writing system compared to Japanese 92–6 ‘In a Station of the Metro’ 96–8 phonographic units 93 ‘time is money’ 95–6 equivalence 46 Evans, D. and A. Helbo 28 Ewe 184 Fairley, I. R. 233 n. 8 Fauconnier, G. and M. Turner 16, 22, 36, 38 Feibleman, J. K. 29 Fenollosa 194, 197, 206, 207 Finegan, E. 196 Firstness, Secondness, Thirdness 29–30, 33 Fish, S. E. 23 Fleischman, S. 191 Fónagy, I. 46, 219 ‘40–Love’ 105–6 metaphor–icon links in 107 Freeman, D. C. 71, 156, 231 n. 5 Freeman, M. 26, 27 Fromkin, V. and R. Rodman 195, 208

Index 255 Fuji, Y. 180 ‘Fuukei (scenery)’ 112–15 Gavins, J. and G. Steen 27 Givón, T. 193 global metaphor 57 in ‘Easter Wings’ 59 in ‘Love’s Philosophy’ 66–7 glottic and non-glottic writing 213 glottic phonographic representation 214 Goody, J. 210 grammatical metaphor 41–5, 108, 117–18 asymmetrical meaning in asymmetrical form 183–4 difference in form 176–9 in ‘Easter Wings’ 61–2 elaborated 175 marked meaning 183–4 relational diagrams 175–9 repetition as 49–50 semantic relevance 180–1 similarity of form 178–9 structural diagrams 179–85 symmetrical meaning in symmetrical form 182–3 temporality 179 typology 173–86 grammaticalisation 17, 184, 188–90 Haberland, H. 208 Haga, J. 199 haiku (form of Japanese poetry) 7, 81 and blending model 71 and cognitive metaphor 70–2 and the cultural cognitive model 72 literature on 17–18 orthographical revisions 118–25 phonological revisions 146–56 subject matter 7 visual presentation 98–100 see also Oku no Hosomichi Haiman, J. 15, 28, 31, 43, 176, 182, 192, 193 Haley, M. C. 33 Halliday, M. A. K. 42, 43, 228 n. 12

Hamano, S. S. 234 n. 6 Hamilton, C. 27 Harada, S.-I. 237 n. 10 Harris, R. 62, 195, 210, 211, 212, 213, 216, 225 Hastings, R. 230 n. 1 Hattori, D. 231 n. 11 Havelock, A. E. 195, 210 Henderson, H. G. 228 n. 8 Henle, P. 57, 229 n. 14 Herbert, George, ‘Easter Wings’ 58–63 hermeneutic viewpoint, metaphors 24 Higgins, D. 230 n. 1 Hinton, L. et al. 28 Hiraga, M. K. 17, 18, 49, 122 Hiraga, M. K. and J. RadwanskaWilliams 28 hiragana (syllabary words of Japanese origin) 10, 18, 93, 94, 198, 199 ‘Hisakata no’ 161–8 consonants in 164 hana and kokoro in 165–7 metaphor–icon links in 168 Hollander, J., ‘Swan and Shadow’ 100–4 Hopper, P. J. 225 horizontal distance 191 icon 30–1, 228 n. 1 index, symbol 30–1 metaphor in 3–5 subtypes of 34 text as 59–63 use of word in connection with computers 3–4 icon–metaphor links, and structural analysis 223 iconic interpretation, navigated by metaphor 10–14 iconic moments, in metaphor 38–41 iconicity cognitive treatment of 27–8 deﬁned 35–6 in discourse 185–6 dominance in metaphor–icon links 4–5

256 Index iconicity – continued in electronic communication 217–18 and kanji 199–207 manifestation in language 15 in metaphor 5–7 and metaphor: clariﬁcation of interrelationship 220–1; as a human issue 227 and non-alphabetic writing systems 196–7 and written language 194–218 icons metaphorical navigation 4 Peircean classiﬁcation 15 resemblance to objects 3–4 Ide, S. 237 n. 10 ideograms 196–7 Chinese characters as 197–9 Ikegami, Y. 18 Ikegami, Y. et al. 17 image, diagram, metaphor 31–4 image content, of metaphor 5–6 images 32 imagic manifestations, visual poetry 100–6 Imoto, N. 17 ‘In a Station of the Metro’ 96–8 indices 30–1 ‘integrational linguistics’ 225 Isami, Y. 225 Ishikawa, M. 185 isomorphic 71 isomorphism 70–89 Jakobson, R. 17, 23, 24, 25, 31, 127, 155, 156, 229 n. 2 on poetic function 43–50 Jakobson, R. and L. R. Waugh 3, 17, 23, 46, 127, 156, 169, 178 Jakobson, R. and S. Rudy 23, 46 Jakobsonian poetics 22, 23 Jakobsonian structural analysis 222 James, V. and E. Jansen 197 Janney, R. W. 229 n. 14, 18 Japanese sound system, compared to English 128–31 Japanese writing system characteristics of character types 93

compared to English 92–6 ‘In a Station of the Metro’ 96–8 phonographic units 93 ‘time is money’ example 95–6 vertical layout 99–100 japonaiserie 115 Jespersen, O. 12, 24, 31 Johnson, M. 16, 25 juxtaposition 8, 13, 110, 160, 204 kaii-moji (compound logographs) 201–3 Kakehi, H. et al. 132, 133, 234 n. 6 Kaneko, Y. 17, 84 kanji (Chinese logographs) 10, 11, 18, 19, 93–4, 116, 198, 199, 223 and the blending model 205–6 categories of 200–1 and iconicity 199–207 iconicity in word formation 204–6 kaii-moji (compound logographs) 201–3 keisei-moji (phonetic–ideographic characters) 200, 202–5 metaphor–icon links in 200–6 as a poetic medium 206–7 shiji-moji (simple logographs) 200–1 shookei-moji (pictograms) 200 katakana (syllabary words for words of foreign origin) 10, 18, 93, 94 Kawamoto, K. 17 keisei-moji (phonetic–ideographic characters) 200, 202–5 Kennedy, X. J. 116 Kikuchi, S. 17 Kobayashi, S. 71 Kress, G. and T. van Leeuwen 216 Ladefoged, P. 128, 129, 234 n. 2 Lakoff, G. 16, 25, 35, 71, 173, 183, 236 n. 7 Lakoff, G. and M. Johnson 15, 15–16, 25, 33, 35, 41, 64, 71, 173, 181 Lakoff, G. and M. Turner 15, 16, 26, 27, 33, 35, 36, 64, 71 Lakoff–Johnson–Turner cognitive model of metaphor 22, 25

Index 257 Landsberg, M. E. 28 Langacker, R. W. 16, 236 n. 4 Langer, S. K. 18, 213, 214, 215 language dynamic nature 224 motivated nature of 220 Languages of Japan, The 186 ‘languaging’ 224–5 Leech, G. N. 131, 132 Lennard, J. 233 n. 8 Li Po 86 linguistic form 174 linguistic iconicity 14–18 studies of 24–5 linguistic poetics 46, 127 linguistic sign, arbitrariness of 14, 24–5 local metaphors 57 in ‘Love’s Philosophy’ 63–5 in ‘Oku no Hosomichi’: autumn poem 81–5; spring poem 75–80 logocentrism 209 logographic writing systems 17 logographs 86, 87 compound logographs 200, 201–3 and phonography, cognitive differentiation between 198–9 simple logographs 200–1 ‘Love’s Philosophy’ 63–70 blending in 66–7 cognitive metaphors 64 contact in 65, 66 fusion in 65, 66 generic spaces 68 global metaphor in 66–7 grammatical parallelism in 63–70 local metaphors in 63–5 metaphor–icon links in 67, 70 opposites in 69 pairing in 65, 68–9 Lyons, J. 195, 208 Ma, Kang Yuen 198 McGough, Roger, ‘40–Love’ 105–6 McLuhan, M. 217, 238 n. 8 McNeill, D. 238 n. 1 marked meaning 183–4

Matsuo, Basho 7, 11, 73, 84, 86, 98, 228 n. 9 Oku no Hosomichi 7–10, 70–89, 118, 119–22, 124–5 ‘Quietness seeping into the rocks, the cicada’s voice’ 146–7 meaning sound as 127–69 vision as 91–126 mental spaces, in blending model 37–8, 52 metaphor dominance, in metaphor–icon links 6–7 metaphor–icon links in ‘The Bells’ 139 cognitive approach 15–17 in concrete poetry 107–8 in ‘Easter Wings’ 60 in ‘40–Love’ 107 in ‘Hisakata no’ 168 iconicity dominance in 4–5 in language 51 in ‘Love’s Philosophy’ 67 in macro-structure 85–9 manifestation: in poems 7–14; prototypical examples 57–90 metaphor dominance in 6–7 in ‘Oku no Hosomichi’, spring poem 79 in polite language 186–92 in ‘The Red Wheelbarrow’ 161 in Shizukasa ya 154 in ‘Swan and Shadow’ 104 metaphorical icon 33 metaphorical juxtaposition 8–10 metaphorical mappings 26 metaphorical transfers 184–5 metaphorical understanding 5 metaphor(s) 32–3 in blending model 37 cognitive theory of 25–6 conduit 41 deﬁned 35–6 diagrammatic content 6 global 57 grammatical 41–5 hermeneutic viewpoint 24 in icon 3–5 iconic interpretation navigated by 10–14

258 Index metaphor(s) – continued iconic moments in 38–41 and iconicity, clariﬁcation of interrelationship 220–1 and iconicity, as a human issue 227 iconicity in 5–7 image content 5–6 local 57 rhetorical viewpoint 23–4 semantic view of 23–4 traditional theories of 23–4 methodology, overview 22–3 metonymy 9, 36, 77, 79, 201 mind, conceptualised as a computer 5–6, 33, 38–9 Miron, M. 209 mirror-image 58, 159, 161 mora, in Japanese and English 128–30, 228 n. 7 Morita, R. 17 motivated nature of language 14, 220 motivation 14–15, 116, 118, 216 motivations, competing 192–3 Müller, W. and O. Fischer 28 Muramatsu, T. 17 ‘My Heart Leaps up When I Behold’ 117–18 Nänny, M. and O. Fischer 28 Niikumi, S., ‘Ame (rain)’ 111–12 Nishida, N. 186 non-alphabetic writing systems and iconicity 196–7 vs. the alphabet 195–7 non-verbal communication 226–7 Norrick, N. R. 185

character choice and meaning of the text 120–2 character choice and meaning of the word 119–20 foregrounding by character choice 123–4 intensifying by character choice 122–3 isomorphism of text structure in 70–89 life and death is a cycle metaphor 87–9 life is a journey metaphor 85–7 revisions 118–25 sound structure 11–13 spring poem 73–80: birds in 77–8; blend in 74–5, 77; local metaphors in 75–80; metaphor–icon links in 79; personiﬁcation in 78–80 structural meaning of character choice 122–5 time is a traveller metaphor 86–7 Olson, D. R. 210 Olson, D. R. and N. Torrance 210 ‘1(a’ 109–11 blended space in 109 oneness 74 Ong, W. 195, 210 onomatopoeia 44–5, 131–3, 154, 234 n. 6 ‘ontological metaphors’ 64 orthographical revisions of haiku 118–25 orthography 119 Owen, Wilfred 27 Ozawa 161, 235 n. 22

Ogata, T. 17 Ohori, T. 184 Oku no Hosomichi 7–10, 233 n. 17 autumn poem 80–5: autumn in 85; blend in 81; futami pun 81, 82, 83; local metaphors in 81–5; metaphor–icon links in 84 bracketing by character choice 124–5

Paradis, M. 199 parallelism 6, 15, 31, 32, 33, 34, 46, 57, 87, 167 grammatical 57 pattern knowledge 5 pattern poetry 58, 106 Peer, W. van 232 n. 1 Peirce, C. S. 29 classiﬁcation on iconography 15 Peircean deﬁnition

Index 259 sign

29–30: Firstness, Secondness, Thirdness 29–30 Peircean semiotics 22 personiﬁcation 77, 78–80, 82 phonemic structures, Japanese and English 130–1 phonetic–ideographic characters 200–5 phonographs, and logographs, cognitive differentiation between 198–9 phonosymbolism 134–56 pictograms 200 Plato, Cratylus 14 Poe, Edgar Allan, ‘The Bells’ 134 poems, manifestation of metaphor–icon links in 7–14 poetic function, Jakobson on 43–50 poetic metaphor, cognitive account 26–7 poetically equivalent 166 poetry, auditory representation in 127–34 polite language distance in 186–7 metaphor–icon links in 186–92 vertical distance 187–91 polysemy 184 Pope, Alexander 127 Posner, R. 174, 180, 181 Poster, M. 211 Pound, Ezra 19, 194, 197, 198, 206, 207 ‘In a Station of the Metro’ 96–8 Selected Poems of Ezra Pound 97 proximity iconicity 180–1 ‘Quietness seeping into the rocks’ 146–7 fusion of voice and silence 151–3 lexical revisions 148–9 metaphor–icon links 154 phonological revisions 149–51 revised versions 147–8 ‘The Red Wheelbarrow’ 157–61 metaphor–icon links 161 mirror-image 158–61 tableau 158

Reddy, M. 41, 174 relational diagrams, grammatical metaphors 175–9 remembrance 4 repetition 111, 115, 185–6 as a grammatical metaphor 49–50 iconic meaning 49 manifestations of 47 role in text 46–8 units of 47 representation, discursive and presentational 18 research issues 224–7 resemblance 4 rhetorical viewpoint 23 rhyme, and rhythm 133–4 Richards, I. A. 24 Ricoeur, P. 23, 24, 229 n. 14 Ross, J. R. 156, 157, 158, 233 n. 15 Sapir, E. 49 Sarup, M. 209 Saussure, F. de 24 Course in General Linguistics 14 Selected Poems of Ezra Pound 97 semantic view of metaphors 23–4 Semino, E. and J. Culpeper 27 Shapiro, K. and R. Beum 12, 127, 131, 132 Shapiro, M. 46, 156 Shelley, Percy Bysshe, ‘Love’s Philosophy’ 63–70 Shibatani, M. 130 The Languages of Japan 186 shiji-moji (simple logographs) 200–1 Shirane, H. 17, 228 n. 8 Shizukasa ya 146–7 shookei-moji (pictograms) 200 sign, Peircean deﬁnition 29–30 sign–object relation 30 signiﬁcation 16 similarity 4, 15, 16, 30, 35, 41, 43, 50, 76, 91, 156, 175, 204 Simone, R. 28 simple logographs 200–1 Solt, R. E. 107 sonorants 12, 13 Sorensen, H. 211, 217

260 Index sound 24–5 diagrammatic structuring 156–68 as meaning 127–69 sound systems, English and Japanese compared 128–31 sound-symbolism 131–3, 209, 234 n. 9, 11 of consonants 132–3 source domain 5, 6 Steen, G. 27 Stockwell, P. 230 n. 1 structural analysis, and icon–metaphor links 223 Summers, J. 230 n. 1 ‘Swan and Shadow’ 100–4 Sweetser, E. 25, 27, 184 syllables, in Japanese and English 128–30, 234 n. 2 symbols 30–1 symmetrical meaning in symmetrical form 182–3

vertical distance 187–91 vision, as meaning 91–126 visual elements 91–2 visual iconicity in poetry diagrammatic icon by grammatical metaphors 117–18 ‘My Heart Leaps up When I Behold’ 117–18 subtle manifestations 116–25 visual poetry 100–16 diagrammatic manifestations 106–15: ‘1(a’ 109–11; ‘Ame (rain)’ 111–12; ‘Fuukei (scenery)’ 112–15 imagic manifestations 100–6: ‘40–Love’ 106–7; ‘Swan and Shadow’ 100–6; metaphor–icon links in 104 vowels in Japanese and English 130–1 symbolic value 24

Talmy, L. 183 tanka 7, 8, 17 Tannen, D. 185 Tanz, C. 209 target domain 5, 6 Taub, S. 229 n. 14 temporal distance 191–2 Tennyson 131 text, as icon 59–63 Tomonori Kino, ‘Hisakata no’ 161–8 Toolan, M. 222, 225 Toyama, S. 71 traditional theories of metaphor 23–4 Tsujimura, T. 186 Tsur, R. 229 n. 3 Turner, M. 16, 22, 26, 27, 35, 36, 38 Turner, M. and G. Fauconnier 16, 22, 36, 204 Tuve, R. 230 n. 1

Watanabe, S. et al. 188 Waugh, L. R. 15, 25, 28, 192 Waugh, L. R. and M. Newﬁeld 28, 178 Weber, J. J. 23 Weiss, J. H. 209 Wellek, R. and A. Warren 131–2, 133 Wescott, R. W. 228 n. 6, 237 n. 1 Westerweel, B. 230 n. 1 Wildman, E. 107 Williams, E. 107 Williams, W. C., ‘The Red Wheelbarrow’ 157–61 word afﬁnities 24 Wordsworth, William, ‘My Heart Leaps up When I Behold’ 117–18 Worth, D. S. 46 writing systems the alphabet vs. non-alphabetic writing systems 195–7 English and Japanese 92–6 and written language 194–5 written language cognitive role 19

Ueno, Y. 17 Ultan, R. 209 Ungerer, F. and H.-J. Schmid verbs of vertical motion, metaphor–icons in 190

16

Index 261 discursive and presentational representation 213–15 glottic and non-glottic writing 213 and iconicity 194–218 integrational deﬁnition 212–13 metaphor and iconicity in writing 215–18 oral–literate equation 210 primacy of speech 207–9 primacy of writing 209 redeﬁnition of 221–2

synthetic view 210–12 and writing systems 194–5 Wyss, E. L. 238 n. 10 Yamamoto, K. 17, 232 n. 16 Yamamura, B. ‘Fuukei (scenery)’ 112–15 Yamanashi, Masa-aki 186 Yasuda, K. 228 n. 8 Yeats, W. B. 27 Yule, G. 196